Professional Documents
Culture Documents
Christopher Pollitt, Geert Bouckaert-Continuity and Change in Public Policy and Management-Edward Elgar Publishing Inc (2011) PDF
Christopher Pollitt, Geert Bouckaert-Continuity and Change in Public Policy and Management-Edward Elgar Publishing Inc (2011) PDF
Christopher Pollitt, Geert Bouckaert-Continuity and Change in Public Policy and Management-Edward Elgar Publishing Inc (2011) PDF
Management
Continuity and
Change in Public
Policy and
Management
Christopher Pollitt
Research Professor in Public Management, Public
Management Institute, Katholieke Universiteit Leuven,
Belgium
Geert Bouckaert
Director, Public Management Institute, Katholieke
Universiteit Leuven and President, European Group for Public
Administration, Belgium
Edward Elgar
Cheltenham, UK • Northampton, MA, USA
© Christopher Pollitt and Geert Bouckaert 2009
Published by
Edward Elgar Publishing Limited
The Lypiatts
15 Lansdown Road
Cheltenham
Glos GL50 2JA
UK
References 198
Index 211
v
Preface
This book sprang originally from a simple source. As we worked together
on earlier academic projects we also talked about our national differences
– of custom, language and, inevitably, public policymaking. Eventually
opportunities presented themselves (with a little help, as usual) and we
were able to tailor at least some of our curiosities into the shape of a joint
academic project. And now, three years later, we have a book that com-
pares the police service with the hospital service, Leuven with Brighton,
Belgium with England, and ‘then’ (1965) with ‘now’ (more than 40 years
later years). It focuses on similarities as well as differences, and on conti-
nuities as well as changes. It examines the national level and the local. It
also interrogates a range of contemporary academic approaches to policy
analysis and international comparison. If we have done our job properly,
readers should be able to get at least a flavour of that original fascination
that prodded us to begin. How can two countries, so close to each other
geographically and historically, be so different? Then (after some more
detailed observation) are they really so different, or is that just a superficial
stereotype?
When would-be authors approach publishers with propositions for aca-
demic books the publishers always ask ‘Who is this for?’ Our answer on
this occasion is, first, advanced students of public management and public
policy, and second, their professors. Both these groups, however, already
have a lot of reading to do, so the obvious further question is why they
should add this book to their existing burdens. Here the answer is a little
more complicated. Obviously Continuity and Change in Public Policy and
Management should be of special interest to those who have a particular
focus on Belgium and England, or on hospitals or the police. However,
we have tried to make its relevance go far beyond these (worthy and
legitimate) special interests. Our aim has been to make the book relevant
for anyone who has a general interest in the study of comparative public
management and policymaking. We have attempted to achieve this broad
relevance by using the particular chosen topics as vehicles with which to
examine the usefulness of some much more general theories and models
of the policy and management. We have tried to make these larger sets of
ideas ‘work’ with a lot of detailed material, and have reflected at length on
the successes and limitations of this application.
vi
Preface vii
Much of the public management and public policy literature falls into
a few well-recognized subsets. There are books on ‘theory’. There are case
studies, often of currently fashionable policies. There are a few histories,
looking back over decades. There are methods texts, which discuss the
properties and limitations of different ways of gathering and manipulat-
ing data. One reason why we hope that the present work will be of wider
usefulness and interest is that it offers all these things together. It engages
with theory, it provides new and original case studies, it spans 40 years
of history and it contains considerable discussion of sources, tools and
methods. Indeed, teachers could use it as a core book for a course that
sought to integrate these different aspects of the subject (we hope they
will).
As we write this preface the mass media are full of claims that we are
on the verge of a new era in public policymaking. We have a global eco-
nomic crisis that seems to be prompting hitherto unthinkable degrees of
government intervention in the business sector. We have a new American
President who has achieved remarkable popularity through an extended,
cleverly nuanced exposition on the theme of ‘change’. We have a recently
enlarged European Union which is attempting to play a larger part as a
global actor, well beyond its existing competencies in trade and agriculture.
In both Belgium and England – our two focal countries – we can witness
high levels of public discontent with contemporary governments, and a
sense that the way policymaking is done itself needs to be transformed.
Much is to be gained, we would suggest, from detailed, longer-term studies
of such apparent turning points and new departures. Continuity and
Change is our attempt to provide such an analysis.
Christopher Pollitt
Geert Bouckaert
viii
1. Theory and method in
comparative studies of
organizational change
1.1 INTRODUCTION
We hope that readers will find each and all of these comparisons intrinsi-
cally interesting, but here at the outset we must warn that none of the
four are straightforward. As we shall see, there are many obstacles and
difficult choices involved in each of these comparisons, and the volumi-
nous literature on public policy and management shows quite clearly
that there is no academic consensus on the best way to make them.
Because of this, the discussion within our book will alternate frequently
between two (mutually dependent) levels: first, that of actually compar-
ing ‘things’ (political systems, hospital financing systems and so on); and
second, ways of making those comparisons – how does one do it? To give
readers a framework for the substantive descriptions and discussions
that come later, this first chapter will concentrate on the ‘how’ question,
and will briefly introduce some of the main issues. Right at the end of
the book (Chapter 9), when the substantive analyses are complete, we
will return to these questions of theory and method, to review what we
have learned.
1
2 Continuity and change in public policy and management
This is not a bad start, but readers should be aware that there is a 40-year-old
debate about the definition of public policy, and that the above is but one
candidate, and a very simple, possibly oversimple, one at that (see also Dye,
1972; Hill and Hupe, 2002). Hogwood and Gunn found at least ten meanings,
including policy as an expression of general purpose, policy as specific pro-
posals, policy as a programme and policy as formal authorization (Hogwood
and Gunn, 1984, pp. 13–19). In some ways it is easier to pronounce upon
what a definition of policy should not say than on what it should. Thus, we
should be careful that our conceptualization of policy does not:
● Assume that policies either clearly succeed or clearly fail. More typi-
cally they will appear to have a spectrum of effects, some intended
and some unintended, some large and many marginal. Furthermore,
some effects may be doubtfully attributed to the policy (that is,
analysts cannot be sure that the policy has actually ‘caused’ these
effects). Other effects may actually result from the policy but be mis-
takenly attributed to other developments.
the term policy is usually considered to apply to something ‘bigger’ than partic-
ular decisions but ‘smaller’ than general social movements (Heclo, 1972, p. 84)
Above all, the modern meaning of the word . . . is that of policy as a rationale, a
manifestation of considered judgement . . . A policy is an attempt to define and
structure a rational basis for action or inaction. (Parsons, 1995, p. 14)
While policy refers to a purposive course of actions, this does not exclude the
possibility that purposes may be defined retrospectively. Policy arises from
a process over time, which may involve both intra- and inter-organizational
relationships. Public policy involves a key, but not exclusive, role for public
agencies. (Hill and Hupe, 2002, p. 4)
Taken together, therefore, these are some of the core features of the
animal ‘public policy’.
We see time not just as unvarying and impersonal ‘clock time’, but as
something which is constructed and used by different societies in differ-
ent ways in different periods (Elias, 1992; Nowotney, 1994; Pollitt, 2008).
Thus attitudes to time themselves vary over time and from culture to
culture. There is, for example, a widespread perception in North America
and Western Europe that the ‘pace of life’, very much including politics
and public policymaking, has accelerated during the past few decades
Theory and method in comparative studies of organizational change 5
(Whipp et al., 2002). There is also said to be a change in the way many of
us regard the future. Forty or 50 years ago it tended to be thought of as a
space in which things would improve – there would be economic growth,
technological progress, social betterment. Politicians held out brave
visions of a better world tomorrow. More recently, however, the future
seems to have become a more troubling and threatening space – populated
with demographic crises, new epidemics, disastrous climate change and
other frightening prospects (Nowotney, 1994).
The period covered by this book stretches for four decades, from the
mid-1960s to 2005. No choice of period is perfect for all analytical pur-
poses, but we would claim that our selection does have some logic.
To begin with, 40 years should be enough to allow us to see some of the
‘longer waves’ in policy development, patterns that are not so visible if one
confines oneself to the common three to five years’ study of some atten-
tion-grabbing contemporary policy and where it immediately came from
(Pollitt, 2008). One leading policy scholar has declared that ‘a number of
recent studies suggest that time periods of twenty to forty years may be
required to obtain a reasonable understanding of the impact of a variety of
socio-economic conditions and the accumulation of scientific knowledge
about a problem’ (Sabatier, 1999, p. 3).
Second, the 1965–2005 period covers a development from high mod-
ernism (with the 1960s in many ways the acme of hopes for rationalist
policymaking) to sceptical postmodernism (with the governments of today
apparently wrestling with ‘wicked’, won’t-be-solved problems, declining
public trust in governments and the dwindling and disappearance of the
grand reform programmes and major ideological controversies which are
said to have characterized the 1960s).
Third, particularly in our local case studies (Chapters 6 and 7), the
40-year span means that we have been able to interview and/or engage
in first-hand correspondence with some of those who were involved.
This gives us an extra source of data – particularly valuable for discover-
ing what expectations and frames of reference were entertained by key
decision-makers at the times that policies shifted. For obvious reasons this
valuable source dries up as one pushes further into the past.
We are well aware that there is a vast literature, spanning public and
private sectors, on ‘change management’. This literature includes lengthy
debates about what counts as change and what does not, and what the
differences are between ordinary, continuous change and ‘discontinuous
6 Continuity and change in public policy and management
change’ (that is, innovation) (Osborne and Brown, 2005, pp. 4–7). In the
last few years this debate has been paralleled by a similar set of definitional
manoeuvres about what is and is not an ‘innovation’ (for example Hartley,
2005). To be candid, we find much of this slightly overdone. The word
‘change’ denotes an exceedingly common, everyday concept, and there
is something faintly absurd about academics spending page after page
trying to turn it into a precise technical term, or rather an entire family
of technical terms. The everyday concept contains several notions – that
of time passing (from before the change, through the change until after
the change) and of one entity or system moving between different states.
Thus change is change in something, and the something endures its change
(‘John is much changed since I last met him’). A country may change its
electoral system from first-past-the-post to proportional representation,
but it is still the same country and it still has an electoral system. Similarly,
the UK in 1948 underwent a huge change in its hospital system, moving
from a mixed system of public and private hospitals to one where almost
all hospitals were owned and managed by the central state. Yet we can still
refer to the UK hospital system, both before and after; we are not talking
about something being entirely replaced with something else from a com-
pletely different category.
What we wish to be able to do is to distinguish different scales of change
in the two public services that we have chosen to study. We want to be able
to say ‘this was a big change, but that was only a small one’, and be able
to justify the distinction. This is not entirely straightforward (although
perhaps not as difficult as some of the change management literature
makes it out to be). Fortunately the policy literature has some helpful sug-
gestions. Think, for example, of an annual budget. It changes all the time,
but usually only incrementally – this programme gets 2 per cent more than
last year, that programme gets 3 per cent less, and so on. A bigger change
may come if there is a fiscal crisis and the government is suddenly obliged
to reduce public spending by 20 per cent over three years – although even
this may be achieved through the usual procedures. But still deeper forms
of change involve altering the procedures, techniques and categories of
the budget – this actually affects the dynamics of the fiscal debate and the
structure of the budget. So the budget categories may be changed from line
items to programmes, and the accounting system from cash to accruals
(Pollitt and Bouckaert, 2004, Chapter 4). These are big changes. Yet there
is an even more profound form of change, and that is when policymak-
ers actually begin to think about the budget in a different way, to see it
as something different from what they thought it was before. Previously,
perhaps, it was thought of in the traditional way, as an annual process
for allocating public expenditure to approved activities. But now there
Theory and method in comparative studies of organizational change 7
● First level of change: changes in the levels set for specific policy
instruments (for example grants for x are increased from 20 to 30
per cent; the percentage of police allocated to anti-terrorist duties is
increased from 5 per cent to 12 per cent).
● Second level of change: shifts in the types of policy instruments actu-
ally used (for example the Conservative government’s introduction
of market mechanisms partly to replace organizational hierarchies
in the National Health Service (NHS) after the 1989 White Paper,
Working for Patients (Department of Health et al., 1989)).
● Third level of change: change in the overall policy paradigm (for
example if some British government were to reconceptualize the
NHS as no longer a universal service, but rather as a residual service
for people who could not afford to purchase private health care
insurance).
task because such professionals normally demand and require large areas
of discretion in which to exercise their skills. What is more, their profes-
sional associations tend to be strong and well organized, as anyone who
has ever had to negotiate with medical associations or police unions will
confirm. These factors have implication for the ways in which new policies
descending from ‘above’ may be responded to.
Both services are also having to cope with rapidly changing technolo-
gies that may have major impacts on how their work is organized and,
indeed, on what work they can actually undertake. For the police the
advent of advanced information and communication technologies (ICTs)
have brought them the ability to check number plates, identities, local
crime trends and many other things almost at the touch of a button. The
patrolling ‘bobby’, locus of a historically high level of discretion and inde-
pendence, used to be only occasionally in touch with ‘the police station’.
Now he or she is in constant, instant touch and is visible on the screens
in the central operations control room. In the investigation of crimes,
forensic technologies such as DNA analysis have revolutionized what can
be discovered and proven to the satisfaction of the courts. In the hospital,
new technologies permit doctors and nurses to diagnose and treat condi-
tions that would have gone unnoticed, or which would have remained
essentially untreatable even a decade or so ago. The implications of this
constant technological advance for budgets, for training, for rationing
decisions and for management itself are enormous.
In addition to coping with changing technologies, both services are
directly affected by changing social habits and norms. Changing diets
and exercise regimes lead to new patterns of disease. The ever-developing
‘consumer society’ and ‘globalized economy’ provoke new patterns of
crime. Both services have had to begin to come to terms with serving
multi-ethnic, multicultural communities, where different social groups
may have differing values and norms in relation to both ‘health’ and
‘crime’. Shifting social norms have also led to far more questioning of
‘the experts’. Both doctors and the police can expect to be questioned
and criticized by their ‘clients’ (let alone the mass media) far more than
would have been usual in the 1950s or even 1960s. Records seem to show
that, in the UK at least, the number of occasions when citizens have been
prepared to use physical violence towards health staff or police officers
has increased markedly.
Both services are labour intensive, and involve a lot of face-to-face inter-
action with citizens. They are largely (though by no means exclusively)
‘people services’. Unlike some parts of the public services, therefore, most
citizens think they know something about hospitals and the police. This
helps to make and keep them in the public and political eye.
10 Continuity and change in public policy and management
Because of their direct implications for our health and safety, and the
sometimes dramatic nature of their interventions, both services have long
been major foci for the purveyors of popular culture. The number of TV
series about hospitals or police services are legion. The TV and newspa-
per coverage of high-profile crimes and health miracles and scandals is
endless. Top doctors and police officers have become regular performers
on the TV news. Thus both services have a strong public image – or rather,
images – and in subtle and important ways these images shift over time
(McLaughlin, 2007; Van den Bulck, 2002; Van den Bulck and Damiaans,
2004).
Thus there are a number of dimensions on which we can compare
the responses of the two services to broadly similar challenges. Whilst it
remains true that they are different from each other in important respects,
they are not so different as to make comparison fanciful or pointless. Their
similarities mean that we can consider them not as ‘most similar’ but as
‘fairly similar’, or similar enough to make differing patterns of develop-
ment thought-provoking and stimulative of further analysis.
In a very general sense, Belgium and England are similar. Both are
advanced industrial economies; both have a heritage of now-much-
declined heavy industry (coal, steel, heavy engineering); both were until
recently colonial powers; both are liberal democracies and, of course, both
are geographically situated in north-west Europe, as close neighbours.
However, the institutional structures of England and Belgium are very
different. The UK is frequently cited as a political system which is unitary,
centralized and majoritarian (Pollitt and Bouckaert, 2004, pp. 292–6). It
tends to produce strong, one-party central governments which are able to
impose their legislative programme on the legislature without too much
difficulty. Belgium, by contrast, is a federal, decentralized and multiparty
polity. Much of its recent political history has centred around a series of
constitutional reforms designed to give greater autonomy to the three
regions (Flanders, Wallonia, Brussels) and three language groups (Dutch,
French, German) (De Winter et al., 2006; Pollitt and Bouckaert, 2004,
pp. 216–20). Getting major reforms through the complex, decentralized,
multiparty Belgian political system is fiendishly difficult (Witte et al., 2000,
pp. 293–7).
Therefore one of our reasons for selecting Belgium and England, in
combination with the hospital service and the police, is that while the two
Theory and method in comparative studies of organizational change 11
services share some important characteristics and the two countries are
neighbours and subject to many of the same socio-economic pressures,
the institutional structures are so different. This provides an interesting
test: if policies are mainly determined by broad socio-economic circum-
stances and/or by the technical characteristics of the service, then we
might expect roughly similar policies to emerge in these two countries.
If, however, institutional frameworks have a significant effect, then the
expectation would be for greater differences between policies in England
and in Belgium. This is rather a crude test, and our analysis is about much
more than just this ‘big picture’ similarity or difference, but one factor
behind the selection of these two countries was our interest in explor-
ing the influence that their hugely contrasting governmental institutions
might have on what kind of policies got made, by whom and with what
effect.
First, one might ask, why address the local level at all? The answer is
that it has been clear from many studies in many countries that national
policies are not simply implemented, formulaically, by regional and local
authorities: they are frequently reinterpreted, modulated, adapted, diluted
or even resisted (for example Hill and Hupe, 2002, pp. 127–33; Pollitt et
al., 1998). Thus, at the very least, a rounded understanding of the policy
process requires attention to local implementation as well as national
promulgation. Beyond this, however, local authorities and organizations
have, in varying degrees, their own powers of policymaking. One central
question for subnational government scholars has long been the degree of
independence or autonomy a given authority or organization possesses
from the national government. Whilst constitutional and legal provisions
tell an important part of this story they usually do not tell it all. Dynamic,
shrewdly led local authorities or local organizations may be able to make
themselves more room for policy maneouvre than unimaginative, slug-
gishly led authorities or organizations, even where both inhabit the same
legal framework. Thus our comparison of Brighton and Leuven enables us
to observe degrees of policy dynamism and degrees of policy autonomy,
and thus contribute to the current debates on multilevel governance
(Bache and Flinders, 2004)
Second, why choose these two particular cities? Sheer convenience was
one reason for selecting Brighton and Leuven. Each of us was well enough
connected in these two cities to be able to gain access to key people and
12 Continuity and change in public policy and management
papers without too much difficulty (and with a lot of generous help).
Convenience, of course, is not academically respectable as a criterion for
social science research design, but it has its advantages nonetheless.
Convenience was not the only factor, however. We were looking for
large towns or small cities that possessed both advanced hospital facilities
and fairly sophisticated police forces. We did not want capital cities, which
tend to have their own particular dynamics and preoccupations, especially
for the police, but also to a significant extent for hospital services. On the
other hand we did not want remote cities or those wrestling with hard
cases of industrial decline – we wanted reasonably prosperous, reasonably
cosmopolitan places that were regular players in national and regional
policies and were not strikingly unusual along some significant dimension.
Brighton and Leuven both fitted these specifications.
England and Belgium are our two country case studies. Brighton and
Leuven provide us with two local case studies. The case study is a very
commonly used approach in the social sciences in general, and in public
policy and public management in particular. Yet its status and purpose
remain contested. At one extreme, case studies may be frowned upon
as one of the weakest research strategies. From this perspective they are
‘soft’, qualitative investigations which are vulnerable to all sorts of biases
and which, at best, can be thought of as ‘small N’ research – a poor cousin
of the ‘large N’ statistical analysis of whole populations of policies or
countries that we should really be pursuing if we want to do serious science
(King et al., 1994). At the other extreme, case studies are seen as a highly
valuable and rather flexible research strategy that can be used for a variety
of important purposes, including testing theories, tracing key causal proc-
esses and mechanisms and ‘congruence analysis’ (Blatter, 2007; Blatter
and Blume, 2007).
One’s perspective on case studies depends to a considerable (though
not absolute) degree on one’s theoretical affinities and orientation. These
orientations are discussed in the following section. Therefore, at this early
point, we will not go any further into the case study debate. Rather we
will proceed with our study, including the elaboration of its theoretical
framework and the delivery of our case studies, and then return, in the
final chapter, to the debate about the potential of this approach. Since we
have chosen to make case studies an important part of our own work it
will not come as a surprise to readers to be forewarned that we find them
very useful.
Theory and method in comparative studies of organizational change 13
First Point
‘History matters’: what happened before often affects what happens later
(Pollitt, 2008; Tilley, 2006). For example, the choice of a particular kind
of constitution has many effects on what can and cannot be done subse-
quently (of course, the constitution can be overthrown or amended, but in
most countries this is not an everyday occurrence, and for most practical
purposes governments and parties can only act as the constitution permits
them to act, more or less). The question ‘What is history?’ has no single,
simple answer (Pollitt, 2008, Chapter 2) but for our purposes it can be
considered as a form of scholarly description and analysis which uses a
narrative form to focus on the sequence of developments over time, and
which has developed specific and meticulous techniques for identifying,
classifying and interpreting various sources of data about the past. History
is not theory-free, but the approach to theory is principally inductive and
inclusive. That is, explanations are produced by constructive attention
to many details and aspects and there is a willingness to try out a range
of competing explanations before selecting the one that seems to do the
best job of synthesis. The synthesized narrative is created by the histo-
rian, not simply revealed as some pre-existing, underlying reality. Unlike
16 Continuity and change in public policy and management
Second Point
Third Point
Strategic action (with all the caveats included in the above quotation)
takes place within institutional contexts but also reshapes those contexts.
As Hay and Wincott suggest, strategic action has both direct effects (for
example direct organizational restructuring to achieve ‘decentralization’)
and learning effects (for example where the consequences of attempted
change makes the actors involved realize that some aspect of their insti-
tutional framework is more or less strong than they realized) (Hay and
Wincott, 1998, p. 956).
Fourth Point
Result of change
Within path/incremental Radical/transformation
Process of Gradual A. Classic B. Gradual, but
change incrementalism eventually
TORTOISE fundamental change
STALACTITE
Abrupt C. ‘Radical D. Sudden, radical
conservatism’ – rapid change
return to previous EARTHQUAKE
ways (punctuation)
BOOMERANG
This is quite a stable situation over time, and much of the classic literature
on incrementalism addresses change of this type (Lindblom, 1959, 1979).
Incremental change is relatively predictable, at least in the sense that the
participants do not usually have to worry about sudden, radical shifts.
One of the best-known examples comes from Wildavsky’s seminal work
on budgeting in ‘advanced’ Western states (Wildavsky, 1986). Let us say
that the English health service gets 2 per cent more in this year’s budget
but the police get 3.5 per cent. However, next year, or the year after, the
health service will get a slightly bigger margin of growth than the police,
and so the relative budget shares of the two services will not change very
much over time.
In the second type (box B) the process of change is again gradual, but
here the direction of change is constant, so the ‘increments’ build up, one
upon the other. One might call it ‘Stalactite’ change, in the sense that,
without anything dramatic happening, eventually water slowly dripping
in the same direction cumulates in a major new feature. It is not always
remembered that Lindblom himself recognized that such a unidirectional
process could eventually lead to ‘a drastic alteration of the status quo’
(Lindblom, 1979, p. 517). In the case of Belgium, for example, it could
be argued that a series of incremental transfers of competencies from the
national level to the regional level eventually reached the point where there
was a de facto federal system.
Box C denotes a pattern of change which we think is only rarely
acknowledged in the policy literature, but which may be rather more
Theory and method in comparative studies of organizational change 19
These, then are four broad possible patterns of change – the BEST
typology, as we will call it, reading the four cells anticlockwise from the
bottom left-hand cell. We will inspect our two countries and two services
to see whether one or more of these categories fits the chronologies that
we will set out. But these are not the only patterns that may interest us. A
number of previous works in the fields of public policy and management
have found policy ‘cycles’ or alternations or other sequences (Pollitt, 2008,
pp. 21–4). What is more, there may be different temporal patterns at dif-
ferent levels in the policy system – for example, cycling between centraliza-
tion and decentralization at the national level while there is simultaneously
some quite significant Earthquake or ‘punctuation’ happening in one or
more specific localities. In the chapters to come we will search for, and
find, a range of patterns, and patterns within patterns.
Fifth Point
Both our histories and our institutions are ‘socially constructed’, in the
sense that their forms depend on the choices of their writers, founders and
reformers, and those choices, in turn, very probably have something to do
with the beliefs and interests of the relevant actors. This means that they
are constantly reconstructed too, which is one reason why social construc-
tivists have difficulty with positivistic treatments of social phenomena that
assume that a ‘parliament’ or an ‘election’ or even a ‘hospital’ or a ‘profes-
sion’ means the same thing over long time periods, or all over the world.
On the contrary, say the social constructivists, scholars should always
pay careful attention to how these concepts are articulated in any given
context (time, place, culture). Even such apparently ‘hard’ management
technologies as performance indicators are best understood as socially
defined (Moynihan, 2008). Furthermore, the patterns we see over time are
themselves socially constructed:
One of the most remarkable features of human memory is our ability to men-
tally transform essentially unstructured series of events into seemingly coherent
historical narratives. We normally view past events as episodes in a story (as is
evident from the fact that the French and Spanish languages have a single word
for both story and history . . .) and it is basically such stories that make these
events historically meaningful. (Zerubavel, 2004, p. 13)
This is not to say that any story is as good as any other, or that factual
accuracy is not a crucial part of the historian’s craft. But it is to say that
there is always more to it than a simple reflection of the ‘facts’. That is one
reason why the histories of particular periods and events are being con-
stantly rewritten and reinterpreted.
Theory and method in comparative studies of organizational change 21
Sixth Point
HIs sometimes have trouble with using their temporal and institutional
patterns as explanations. As many but not all HIs have realized, they need
something else, ‘behind’ or ‘beneath’ the patterns to actually power an
explanation. Thus, for example, to say that something is ‘path-dependent’
is not to explain why it remains fairly stable over time. It merely describes
a pattern of development – stable, moving down the same, constant
track. To explain why this happens requires the theorist to spell out some
process(es) or mechanism(s) which holds things in this steady direction
(Kay, 2006; Pierson, 2004). Such processes may be difficult to observe,
and/or there may be so many candidate processes that it is hard to decide
which to include in the explanation and which to leave out. Some theo-
rists have attempted to classify these processes, but there is no generally
accepted categorization in use (Pollitt, 2008, pp. 43–4). Whilst some theo-
rists have taken an essentially economistic approach (for example Arthur,
1994, identifies large sunk costs, learning effects and network effects) our
own belief is that a broader spectrum needs to be considered. Thus cultural
and political mechanisms need to be added to those which operate through
the assumptions of economic rationalism. At all events, identifying and
evaluating key processes of positive and negative feedback to policymak-
ers and implementers is a crucial part of the construction of a convincing
and well-grounded analysis of policy change. This point connects strongly
with the third point (above). Actors with strategies are often those who,
by understanding underlying processes, are able to seize upon ‘windows of
opportunity’ for significant change (Kingdon, 1995).
We have now spent enough time on the sketch map. It is time to start
the journey.
2. National reforms: the Belgian and
English regimes
2.1 BRITAIN AND BELGIUM: SIMILARITIES AND
DIFFERENCES
22
National reforms: the Belgian and English regimes 23
To some extent there has been a softer, more nuanced echo of this
stereotype in the academic literature on public policy and management.
To begin with, Belgium is often completely absent from comparative
texts – it seems to appear much less often than other small countries such
as Denmark or Sweden or the Netherlands. In addition, when it does
appear, it is usually labelled as a slow-mover, a polity which is so institu-
tionally complex and so locked into elaborate consensualist processes by
its internal linguistic divisions that it finds it hard to act dynamically or
decisively. (For a range of internal views by Belgian scholars, see the very
useful November 2006 special issue of West European Politics, including
article titles such as ‘Does Belgium (still) exist?’ Billiet et al., 2006). Thus,
in mainstream comparative politics texts, Belgium is (or at least used to
be) deemed to be highly consensualist and corporatist, almost in the oppo-
site corner from the UK in most diagrammatic representations (Lijphart,
1984, 1999). And in comparative public management texts we see the same
thing: in general Belgium comes over as slow-moving, incrementalist and
conservative, whereas the UK has been one of the most consistently radical
‘core New Public Management’ states (Pollitt and Bouckaert, 2004).
More recently, however, a different perspective has appeared. From this
alternative point of view Belgium appears relatively stable in policy terms,
whereas the UK can be seen to be in the grip of ‘hyper-modernism’ and
‘hyper-innovation’ – a frenetic urge to interfere and reform which carries
many costs as well as benefits (Moran, 2003; Pollitt, 2007, 2008). A third
view might be that, while Belgium has been slow-moving with respect
to policymaking, it has witnessed a surprising amount of constitutional
reform and change in the system of political parties over the 40 years of
our study (Deschouer, 2006; De Winter et al., 2006).
There is no short answer to these stereotypes. Through the remainder
of this book, beginning in this chapter, we will begin to unpick the various
elements, accepting some, qualifying others and reframing or rejecting the
remainder.
civil servants are party members. Since the late 1990s efforts have been
made to break away from the cycle of political appointments at the highest
levels, and to give civil servants both a bigger policymaking role and more
room to manage in a professional way, but progress has been slow. The
biggest reform at the federal level – the ‘Copernicus’ initiative from 2000 –
made some significant changes, but the original intention to transfer much
policy formulation from ministerial cabinets to more professional policy
units staffed by civil servants was soon dropped (Brans et al., 2006). In its
2007 review of the Belgian system of public management the Organisation
for Economic Co-operation and Development (OECD) concluded that:
In short the (many) political parties remain the kernel of the Belgian
system, controlling both policymaking and many public appointments.
They have their own ways of evading bureaucratic congestion:
UK Belgium
Average effective number of political parties represented 2.11 4.32
in the lower houses of parliaments following elections,
1945–961
Proportion of time when cabinet was single party, 1945–96 100% 8.3%
Index of executive dominance2 5.52 1.98
Notes:
1. This is not the same as the number of self-defined parties. For details of measurement
see Lijphart (1999), pp. 74–7.
2. Lijphart’s Index of Executive Dominance is a complex measurement, and its full
derivation is too complex to spell out here. Essentially it is related to two different
measures of the average durability of cabinets in each country, where longer duration is
taken as a sign of a strong executive, and produces a higher index number. For details
see Lijphart (1999), pp. 129–39.
and 1990s. The gentlemanly club culture was the stagnating residue of
a nineteenth-century political and social settlement, a matter of infor-
mal, semi-private coordination between elites. The new regulatory state
emerged from the political, economic and social crises of the 1970s. It was
marked by a hyperactive process of institutional upheaval, with organiza-
tional restructurings, remodelling of the public service professions and the
vigorous creation of formal frameworks of targets and criteria by which
to steer other social actors. Whether or not one accepts the full thrust of
Moran’s analysis it is undeniably the case that, first, there have been many,
many reforms and, second, that the informal and even amateur style of
the senior civil service of the 1950s and 1960s has been replaced by a much
more formal and managerial image (Bovaird and Russell, 2007; Pollitt and
Bouckaert, 2004, pp. 292–9).
Some of the differences in political systems can be represented in tabular
form. Table 2.1, extracted from Lijphart (1999), shows some important
dimensions. Table 2.2 summarizes some of the key differences in the
nature of executive government.
with the first (1965–85). However, that acceleration appears to have been
far greater in England, with respect both to the police and the hospital
system. Table 2.3 sets out the barest bones of the story, to be filled out in
subsequent chapters.
The table already hints at, but does not fully register, the greater degree
of ‘activism’ in English policymaking. For the police, for example, in the
Belgian case one could refer to ‘the’ (singular) ‘police reform’ and most
people would understand that you were referring to the police reorganiza-
tion of 1998–2002. But if you mentioned ‘the police reform’ in England it
would make no sense, and your respondent would almost certainly reply
by asking which reform you meant.
It was not only that more measures and decisions were taken in England,
it was that many of these were more aggressively implemented than in
Belgium. For example, by the late 1990s, performance indicator systems
in both the English police and the English hospital system had real ‘bite’,
in the sense that senior managers could be (and sometimes were) dismissed
for missing targets, and performance failures led to rapid, intrusive central
intervention. By contrast the Belgian hospitals had no national perform-
ance measurement system, although indicative targets had been formulated
for certain dimensions (for example average lengths of stay). Certainly the
28 Continuity and change in public policy and management
financial crisis in many municipalities when they became liable for the defi-
cits of public OCMW/CPAS (Openbaar Centrum voor Maatschappelijk
Welzijn/Centre Public d’Aide Sociale) hospitals. So, despite starting
‘further back’ in terms of control, the Belgian government tried various
forms of indicative planning and norm-and-quota and limit setting to
restrain and guide the expansion of hospitals.
There were also some common elements to the evolution of the police
services in the two countries, despite very different organizational histories
and frameworks. In Belgium, as in England, our 40-year period saw the
police become a more specialized, high-tech operation. Police cars made
the first impact on the traditional ‘bobby on the beat’, and they were fol-
lowed by computers, mobile phones, closed-circuit television (CCTV),
DNA testing and many other developments. It is not much use sending
a generalist police constable out to deal with Internet crime, complicated
fraud, budding terrorist cells or cross-border people trafficking. Officers
have to be given specialist training for such assignments, as increasingly
they also need to receive training for victim support, crime scene manage-
ment and riot control. Furthermore, as in many other countries, the forms
that police filled in whenever they recorded a crime became more elaborate
and formulaic and ‘feeding the database’ became a more prominent part
of everyday police activities (Ericson and Haggerty, 2002).
Yet at the same time both countries experienced strong political and
popular demands for more emphasis on ‘community policing’ that would
be responsive to the particularities of specific communities and localities.
Later on we may see how difficult success in this area is to achieve, but
for the moment we may simply note that this was a recurrent response to
violent street crime and local breakdowns of public order (especially where
the latter had an anti-police, ethnic minority dimension, as they did on a
number of occasions in England).
In the next two chapters, maintaining our comparative focus, we will
extend this analysis by moving deeper into the specifics of each policy
sector.
3. National reforms: hospitals
3.1 INTRODUCTION
The main players in England over the whole period were central govern-
ment, the medical profession (both locally, in the hospitals, and centrally,
through its national representative associations) and the hospitals them-
selves. However, the balance between these three has certainly shifted more
than once, and arguably an important new player has emerged since the
mid-1980s, in the shape of professional National Health Service (NHS)
managers. Local governments have never had any serious influence over
the development of NHS hospitals, except insofar as they can limit infra-
structural developments through their use of land use planning legislation.
An authoritative recent textbook divides the key developments in formal
organization into the following phases (Harrison and McDonald, 2008):
32
National reforms: hospitals 33
The authors do not assign a title to the most recent period, since the arrival
of the New Labour government in power in 1997. In keeping with the
maelstrom of reforms since then we are tempted to call it:
mechanisms, but within five years it was moving back towards them
again. Throughout this period there was a hectic rate of institutional
change, with countless new organizations being invented at local,
regional and national levels, on the purchasing ‘side’, the provider
‘side’ and especially in the shape of new national regulatory agen-
cies (see Pollitt, 2007 for details). The role of managers in all this
was further strengthened and they became, in the words of one
professor of healthcare policy ‘“change junkies” – able to deliver
whatever structures or systems their political masters and mistresses
demand’ (Hunter, 2006, p. 209). In 1998, for the first time, trust chief
executives (most of whom were managers with no clinical training)
became formally responsible for the clinical as well as the financial
performance of their organizations. Doctors, meanwhile, became
subject to tighter and tighter guidance and regulation in their prac-
tice of medicine – mandatory clinical audit, retraining programmes,
a proliferation of guidelines and protocols and the appearance of a
National Institute for Clinical Excellence (NICE) which appraised
the effectiveness and cost-effectiveness of specific clinical interven-
tions and then gave authoritative advice to the NHS on what should
and should not be available.
The main characteristics of the Belgian health care system date from deci-
sions taken after the Second World War, when a compulsory public health
insurance system was set up, based on the principles of:
Department of Department of
Public Health and Social Affairs
the Environment
Regulation
National Institute
for Sickness and
Invalidity Insurance
Transfers
Mutualities
Source: Ministry of Social Affairs, Public Health and the Environment – see European
Observatory on Health Care Systems (2000), p. 10.
medical activity (for example nursing, hotel costs) and one for medical
services. The latter are covered by a fee-for-service system, the former by
a prospective budget based on per diem and patient day quotas. Hospitals
get a share of the fees earned by the resident doctors. Hospitals receive
most of their money from the federal level, but are managed by the com-
munities. This can lead to tensions, as when the communities award higher
salary increases than the federal authorities have allowed for.
One central thread in the story of English hospitals throughout the period
is the drive to rationalize. At its inception in 1948 the NHS inherited an
unplanned patchwork of hospitals of all shapes and sizes. For the first 15
years there was very little new building, but from the mid-1960s onwards
there were a series of attempts to close down small, old or inconveniently
placed hospitals and replace them with a more planned, efficient system.
Later, as medical techniques advanced, average lengths of stay began to
fall, and conditions which had once required inpatient treatment could
now be handled by primary care doctors or on an outpatient basis.
Together with an ongoing (and periodically intense) government pressure
for economies, this led to a marked fall in the total number of beds in the
system (see Table 3.1).
The number of separate sites offering hospital services also declined –
from 2063 in 1978 to 1624 in 1990/91 (surprisingly, figures for sites are not
available since then, because the statistics are collected by trust, and many
trusts have more than one site – Hensher and Edwards, 1999). However,
the falling numbers of both sites and beds have been matched by falls in
the average lengths of stay that patients make. These fell continuously over
our period (from 49.3 days at the start of the NHS in 1949 to 19.8 days in
1979, and much more again since then). Yet activity has increased:
Alongside the growth in day cases, one of the critical mechanisms by which
throughput has been increased and total inpatient activity expanded has been
the fact that the length of stay has consistently fallen at a faster rate than the
number of beds. (Hensher and Edwards, 1999, p. 912)
Note: These figures represent the average daily number of available beds.
1962
Central government published its Hospital Plan for England and Wales
(Ministry of Health, 1962). This marked the first major programme of
new hospital building since the setting up of the NHS in 1948. It also
ushered in a model of the District General Hospital (DGH) which was
to have 600–800 beds and would offer a comprehensive service, including
outpatients and diagnostics (Allen, 1979). At this time the NHS as a whole
was still organized on the ‘tripartite’ basis which had been adopted at its
foundation. Hospitals were the third of the three legs, the other two being
general practitioners and local governments (who provided public health
and prevention, child welfare and various other services). The hospitals
had their own hierarchy, based on 20 hospital regions, each with a medical
school and each guided by an appointed, part-time Regional Hospital
Board (Harrison, 1988, pp. 9–12).
1967
A joint committee of the Ministry of Health and the medical profes-
sion published the first of a series of reports on hospital management
(known as the first ‘Cogwheel Report’). It recommended the setting-up
of speciality-based divisions within hospitals. Each speciality would send
a representative to a medical executive committee, the chair of which
would be the chief spokesman for the profession within the hospital. This
model was widely, though far from universally, adopted (Harrison, 1988,
p. 14).
1974
The first major restructuring of the NHS since its creation in 1948.
Henceforth there were to be 14 Regional Health Authorities and 90
National reforms: hospitals 39
1976
Fiscal crisis. The NHS suffered from the harsh expenditure cuts intro-
duced by the Treasury for the whole of the public sector.
1978
Introduction of a new system for resource allocation between health
authorities, based on standardized mortality rates rather than on historic
costs (the result of the 1976 report from the Resource Allocation Working
Party – RAWP). This led to significant redistribution between geographi-
cal areas (for example towards poorer areas in northern England and away
from wealthier, healthier areas in the south). Given that the redistribution
took place during a period of severe fiscal restraint this meant that some
health authorities actually had to lose money. But this was still very much
part of the era of policymaking-as-planning.
1979
Report from the Royal Commission on the National Health Service (the
‘Merrison Report’): no striking changes, but a large number of piecemeal
recommendations. The government was able to pick and choose what it
implemented.
1980s
During the 1980s there was only one year (1987/88) when there was a
substantial increase in real resources for the NHS, despite evidence that
demographic and technological change required an increase of perhaps
2 per cent per annum simply to maintain the status quo in service levels
(Harrison et al., Chapter 2). Thus the background during the 1980s was
that of an almost constant resource squeeze.
40 Continuity and change in public policy and management
1982
In January the Secretary of State (Norman Fowler) announced measures
to ‘improve accountability’, the main ones being a review process and a
set of performance indicators. Each year ministers were to review the per-
formance of each Regional Health Authority (RHA), and each District
Health Authority (DHA) was to be reviewed by its RHA. Performance
indicators would be developed to assist in these review processes. The 1974
structure was changed.
1983
January: the Secretary of State introduced central control of NHS
manpower.
September: publication of the first national set of NHS performance
indicators (Pollitt, 1985). In the same month compulsory competitive ten-
dering was introduced for NHS laundry, domestic and catering services.
October: the NHS Management Inquiry (the ‘Griffiths Report’) strongly
criticized NHS management and proposed the creation of general manag-
ers for every NHS unit, district and area. At the national level an NHS
Management Board was to be created, supervised by a Health Services
Supervisory Board chaired by the Secretary of State. The inquiry had been
launched by ministers and its personnel consisted of four businessmen,
led by a supermarket executive, Roy Griffiths. Its recommendations were
rapidly accepted by ministers and promulgated as government policy. A
flavour of the report can be gained from the following:
there is no driving force seeking and accepting direct and personal responsibil-
ity for developing management plans, securing their implementation and moni-
toring actual achievement. (p. 12)
1989
The White Paper Working for Patients (Department of Health et al.,
1989) introduced wholesale restructuring of the NHS, along the lines of
an ‘internal market’, with hospitals as providers and health authorities as
purchasers of services on behalf of defined resident populations. At the
heart of this was the creation of NHS trust status, which hospitals could
apply for and which brought them corporate legal status and enhanced
management autonomy. This huge and quite radical change was very
much a ‘bright idea’ developed in secret within a very small group of
politicians and political advisers. It also involved changes in notions of
representativeness, with trusts being run by relatively small corporate
boards which lacked the local authority and trade union members that had
characterized their predecessors.
1998
Hospital trust chief executives (who for the most part had no clinical train-
ing) were for the first time given responsibility for the clinical as well as the
financial performance of their organizations. Trusts were given a statutory
duty to deliver high-quality care.
1999
The government set up the National Institute for Clinical Excellence
(NICE), a powerful regulatory body which had the authority to approve
particular treatments as cost-effective, promulgate clinical guidelines and
approve models of clinical audit. This therefore represented a major incur-
sion into the territory of ‘clinical freedom’.
In the same year another major regulator was created, the Commission
for Health Improvement (CHI). The CHI was to carry out regular reviews
42 Continuity and change in public policy and management
2001
The eight NHS Regional Offices were restructured into four Regional
Directorates of Health and Social Care. In 2003 these were themselves
replaced by 28 Strategic Health Authorities. Thus the organizational
level ‘above’ the hospitals seemed to be in a state of constant flux (see
Pollitt, 2007). Meanwhile, for individual hospital trusts, the government
introduced what became known as the ‘star system’. Each hospital was
awarded three, two one or zero stars according to its composite score
against a range of performance indicators. These were much publicized.
At the same time, and connected with these league tables, the rate at which
NHS chief executives were moved (or fired) accelerated to a high level (see
Table 3.2). The changing number of trusts (due mainly to mergers and new
creations) should also be noticed.
2002
The government announced that it would move towards a system by
which hospital services would be commissioned by Primary Care Trusts
from a wider range of NHS and private hospitals, including envisaged
‘independent treatment centres’. This was, therefore, a return to competi-
tion as a major selection mechanism, and one explicitly aimed at creating
a ‘mixed market’. The government also announced a new type of NHS
hospital, the Foundation Trust (FT). FTs would be given greater financial
1945
The main characteristics of the Belgian health care system date from deci-
sions taken after the Second World War, when a compulsory public health
insurance system was set up, based on the principles of:
1963
August: ‘Leburton’s Law’ created the National Institute for Sickness
and Invalidity Insurance (INAMI/RIZIV). This separated the insurance
44 Continuity and change in public policy and management
system for the health care system from that for the invalidity (incapac-
ity to work) system. INAMI/RIZIV oversees the general organization of
compulsory health insurance but the actual provision of this insurance
falls to statutory sickness funds, the mutualities, which are private non-
profit organizations. As mentioned earlier, they are distinguished by reli-
gious or political affiliations (for example the National Union of Socialist
Mutualities, the National Alliance of Christian Mutualities).
1963
Hospital Act, December: (the ‘Cluster Law’). The first law explicitly
directed at hospitals, which no longer linked hospitals to the compulsory
health and invalidity insurance system. Hospitals were obliged to guaran-
tee hygiene, safety, comfort and quality of care. Cluster’s Law had four
objectives:
1966
The first hospital plan was formulated, with target figures for each
region.
1960s
During the 1960s health care insurance was progressively extended to
wider social categories.
1973
Hospital Act (6 July): this introduced a model of imperative planning. No
new investment or subsidy or approval would be forthcoming unless it
fitted with the federal hospital plan. It was aimed at structural rationaliza-
tion and cost reduction by restricting the supply side (Peers, 1994b, p. 20;
see Table 3.3)
1980
Many elements of health care were delegated to the two communities, and
the ministries of Public Health and Environment were merged.
1982
The Royal Decree of 22 July imposed a Moratorium on the opening of
new hospital beds. Beds had increased by 30 per cent between 1971 and
National reforms: hospitals 45
Year Total beds Average per hospital Beds per 1000 inhabitants
1977 36 366 155 6.63
1980 37 564 152 6.67
1985 37 313 179 6.57
1990 32 596 192 5.65
1982. According to the planning criteria there was a 20 per cent surplus
(11 200 beds)
1986
April: Royal Decree No. 407 introduced a process-led policy: input crite-
ria (bed approvals) were replaced by process criteria (for example numbers
of babies delivered in a maternity hospital).
Explicit in this law was a new understanding of what a hospital was:
1989
Royal Decree on regrouping and merging hospitals.
example endoscopy) meant that patients did not have to spend so long in
hospital. A greater range of care could be given on an ambulatory basis.
However, this carried a number of other consequences, including the need
for hospitals to work more closely with primary care doctors and commu-
nity-based care service, and the fact that the patients that were admitted to
hospital now tended to be sicker and to need more intensive forms of care
(Peers, 1994b, pp. 22–3).
1990
Legislation strengthened state intervention in the process of forming con-
tracts and agreements. If budgetary limits for each care sector are exceeded
fee levels can be reduced. If committees fail to implement the correction
mechanisms the Minister of Social Affairs can intervene to impose them.
1990
November: a Ministerial Decree coupled hospital budgets paid by INAMI/
RIZIV to aspects of hospital structure and activity profile. Target lengths-
of-stay were introduced.
The establishment of Crossroads Bank for Social Security. This ‘bank’
aims ‘to be the motor of e-government in the social sector’ (www.ksz.fgov.
be/En/CBSS.htm, accessed 11 December 2008). It connects up informa-
tion flows between more than 2000 separate institutions, including health
care providers and insurers, greatly reducing the number of different forms
that have to be filled in and the time taken for transactions. In 2006 it won
a United Nations Public Service award.
1993
1 October: a Royal Decree increased co-payments and co-insurances for
visits to generalist and specialist doctors (for specialists from 25 per cent
to 40 per cent of the agreed fee). On 1 January 1994 co-payments and co-
insurances for clinical tests and medical imaging were also increased. The
aim was to reduce system costs by promoting more ‘responsibility’ among
patients. But Belgians can take out private insurance to cover these co-
payments. By 1999 about 30 per cent of Belgians had such insurance.
13 December: national accreditation of doctors was introduced by a
national agreement between the mutualities and the physicians. Fees-for-
service were increased for accredited doctors.
1994
Legislation merged the financial management of all social security sectors,
with the aim of making the social security system independent of state
intervention. This was part of the effort to control the public expenditure
National reforms: hospitals 47
deficit. The background to this and other moves was Belgium’s need to
meet the Maastricht criteria for entry to the European Single Currency.
A Royal Decree of 12 August ‘responsibilized’ the mutualities’ health-
care expenditure. From 1995 the mutualities would receive a prospective
budget to finance the health care costs of their members but they would
be responsible for financing any discrepancy between this prospective and
the actual spending. Initially only 10 per cent of their funds were allocated
prospectively, but by 2000/2001 it was 30 per cent.
1995
The Ministry of Public Health and Environment (see 1980) was merged
with the Ministry of Social Provision, to become the Ministry of Social
Affairs, Public Health and Environment. Yet the responsibility for health
care remains divided at federal level because the separate Ministry of
Social Affairs is still responsible for the National Office of Social Security
(which collects social security contributions) and the National Institute
for Sickness and Invalidity Insurance (which manages compulsory health
insurance).
There were 191 private (mainly non-profit) and 97 public hospitals.
1996
Federal government introduced a Diagnostic-Related Group (DRG)
system for hospital pharmaceutical expenditure (which had been rising
rapidly).
1997
The percentage of total healthcare spending going to inpatient care
reached 39.4 per cent, having risen fairly consistently from its 1970 share
of 25.7 per cent (1980 = 33.1 per cent).
First moves were made to set quotas for the numbers of doctors who
could be accredited, aimed at curbing oversupply.
1998
Hospitals were not meeting basic infrastructural standards by 1 October
1997 were to be closed.
2000
At the turn of the century Belgium had the lowest rate of private health
care expenditure in the Organisation for Economic Co-operation and
Development (OECD) (12 per cent). Total health spending at constant
prices had risen 1970–90, but declined from 1994. As a percentage of gross
domestic product (GDP) health care spending was 7.6 per cent in 1998,
48 Continuity and change in public policy and management
close to the OECD average of 7.9 per cent and considerably lower than the
share in Germany or France.
Whilst both the British and the Belgian governments have, perforce,
needed to take action to control the seemingly incessant growth in the
costs of their hospital systems, their means for doing this have differed
considerably. In the Belgian system, where hospitals are mainly autono-
mous actors rather than state organizations, the government has necessar-
ily acted ‘at a distance’ by attempting to control reimbursement rates and
planning permissions. Only recently, cautiously, and to a limited degree
has government tried to reach inside hospitals to influence what doctors –
the group that, de facto, actually commits a large part of the expenditure –
actually do (for example through doctor quotas from 1997, or threatening
to close establishments which do not meet basic physical standards). By
contrast we can see a major shift in the strategic approach of British gov-
ernments from the middle of the 1980s. Whereas in 1988 Harrison could
write that: ‘In the period from 1948 to about 1982, governments have had
little real interest in control of doctors by NHS managers’ (Harrison, 1988,
p. 22), this statement could not be applied to the period since then. Since
the mid-1980s we have seen governments of both major parties launch-
ing a series of policies which have been intended to strengthen the grip of
professional managers over what happens in hospitals. And – particularly
since the creation of NICE in 1999 – government has attempted directly
to prescribe the content of clinical practice: what protocols should be fol-
lowed for particular diagnoses, what drugs should prescribed, and so on.
This Whitehall incursion into the heart of the clinical arena has not
National reforms: hospitals 49
been instead of, but in addition to, strong financial controls. The central-
ized, state-owned, tax-based nature of the English hospital system means
that the Treasury is able to control the overall size of the NHS capital and
revenue budgets far more directly than would be possible for the Belgian
federal government. Overall, then, we find that the English hospital system
is far more centralized and more tightly government-controlled than its
Belgian counterpart.
But while battles over money and doctors’ autonomy may have claimed
the political headlines, other equally important changes have been running
without so much media attention. Hospitals have become, on average,
bigger and more productive – and this has been supported by government,
in both countries. ‘The current trend in many countries is towards larger
hospitals because average costs are believed to decline with size and because
patient outcomes are believed to be better in hospitals in which clinicians
have the experience of a larger volume of cases’ (Eeckloo et al., 2004, p.
2). This has been a Stalactite-type process, by which the average hospital
of 2005 was much bigger than its average 1965 counterpart. Patients come
and go more quickly and, on average, the inpatients have become sicker
(because lesser conditions can now be effectively dealt with in community-
based facilities or on an outpatient basis). And the technology inside the
hospital’s doors has been transformed, with all sorts of consequences for
staffing and training and financial outlays. In both countries governments
have felt obliged to try to control the spread of the most expensive items of
new medical technologies, both to limit expenditure and to try to ensure a
balanced geographical spread.
Thus we see both strong contrasts between the two systems and also
some very similar underlying trends. One might say that although the gov-
ernments operated with very different political constraints and somewhat
different policy toolkits, they were nevertheless obliged to address some
rather similar-looking problems.
4. National reforms: police
4.1 KEY INSTITUTIONS IN POLICE POLICY
In both countries the key institutions evolved over the period of study,
and some new ones emerged. The most obvious pattern in England
was the creation (or enhancement) of a series of national-level bodies
to deal with the perceived gap left by the famous absence of a national
police force. These included bodies devoted to scrutiny or improve-
ment, such as a strengthened and more inquisitive HM Inspectorate of
Constabulary, the Audit Commission (which has carried out a number
of influential performance audits of different aspects of policing – see, for
example, Audit Commission, 1990, 1996) and, more recently, the National
Policing Improvement Agency. Alongside these have appeared a number
of new executive and operational units, including the National DNA
Database, the National Criminal Intelligence Service, the Home Office
Police Standards Unit and the Serious Organised Crime Agency. Perhaps
most important of all has been the development of the Association of
Chief Police Officers (ACPO):
The Home Office has . . . encouraged (ACPO) to develop a much higher profile
and expand its role, as a means of enhancing the standardization and centrali-
zation of policing. (Reiner, 2000, p. 192; see also Jones, 2008, pp. 7–8)
50
National reforms: police 51
The first ten years or so after the Second World War had been a kind of
‘golden age’ for the police, in the sense that they had been widely regarded
as resourceful, incorruptible and a core part of British society, exemplify-
ing what many liked to believe were its defining qualities of quiet, prag-
matic good sense. They were ‘one of us’ (McLaughlin, 2007, Chapter 1).
However:
During the mid-1970s the fraying of trust was accelerated by a series of cor-
ruption scandals in the Metropolitan Police. Robert Mark was brought in as
an ‘outsider’ Commissioner to clean up the force (Reiner, 2000, pp. 62–4).
1964
The Police Act merged smaller forces to create 49 larger police forces.
52 Continuity and change in public policy and management
1967
‘Unit beat policing’ (UBP) was introduced, leading to more car patrolling
and fewer foot patrols.
1969
The Police National Computer was set up.
1972
The National Reporting Centre was set up.
1976
The Police Act set up the Police Complaints Board.
1977
The Fisher Report found that police had repeatedly violated the rights
of three teenage boys charged with murder. This was one of a number of
incidents that eventually led to the 1984 Police and Criminal Evidence Act
(see below).
1978
The Edmund-Davies report recommended large increases in police pay
and an inflation-proof formula. This was immediately implemented when
the Conservative government came to office the following year and (at a
time of high unemployment) helped make the police an attractive career.
The percentage of graduates applying for police work climbed rapidly,
with later consequences for the generation of individuals who became
chief constables in the late 1990s and 2000s.
1980
The Parliamentary Select Committee for Home Affairs reported that there
had been a growing number of deaths in police custody (from 8 in 1970 to
48 in 1978). This seemed to be yet another area in which the police were
not following appropriate rules and procedures.
1981
After a sweeping Labour victory in the Greater London Council (GLC)
elections the GLC established a Police Committee with a support unit to
monitor police policy. A number of London boroughs also established
‘monitoring groups’. This was generally depicted as ‘politicization’, and
at first there were many tensions between chief constables and these com-
mittees. This is often seen as the beginning of a new, ‘bottom-up’ policy in
which democratically elected local authorities (or at least some of them)
National reforms: police 53
attempted to gain greater scrutiny and control over the police forces in
their areas. (The municipal Belgian police were already under detailed
political control – there was little of the British tradition of ‘operational
independence’ for chief constables.)
The Scarman Report (Scarman, 1981) followed major riots against the
police in a deprived, extensively black area of south London. The subse-
quent Scarman Report ‘dominated police reform throughout the 1980s’
(Reiner, 2000, p. 204; see also McLaughlin, 2007, pp. 66–67, 144–7). It
made many recommendations aimed at improving police–community
relations, and reducing racism within the police.
1982
Sir Kenneth Newman became Commissioner for the Metropolitan Police
and introduced a Scarman-derived strategy that became the model for
many other forces. He based his measures on what he termed a ‘notional
social contract’.
1983
Home Office Circular 114 introduced an emphasis on value for money
(VFM) and required the police to set objectives and priorities. This
marked the beginning of a growing concern with measured efficiency.
1984
The Police and Criminal Evidence Act (PACE) attempted a codification
of the powers of police to investigate crime, and of the safeguards over
the exercise of these powers. It also enabled the establishment of police–
community consultative groups. The powers given by the Act were framed
by the accompanying five Codes of Practice, which ‘provided detailed
procedures regulating stop and search; search and seizure; detention
and questioning of suspects; identification parades; and tape-recording
of interviews’ (Reiner, 2000, p. 177). Failure to comply with the codes
was a disciplinary offence, and a breach could be admitted in evidence in
criminal proceedings. PACE is generally seen as a watershed in making the
police more accountable and professional.
1984–85
The Miners’ Strike. Very large-scale police operations were used during
this highly political strike. The police could not avoid being drawn into
controversy, and they became a target for growing left-wing criticism of
heavy-handed tactics and breaches of civil liberties. For example, in 1984
the (Labour-led) South Yorkshire Police Authority attempted to instruct
its Chief Constable to disband certain units, ostensibly on financial
54 Continuity and change in public policy and management
grounds. But the Home Secretary warned them that they might be in con-
travention of the Police Act 1964, and the Police Authority soon backed
down.
1985
The Prosecution of Offences Act created a separate Crown Prosecution
Service (CPS), removing this responsibility from the police.
Highly publicized urban disorder occurred in Handsworth, Brixton and
Broadwater Farm (all areas with a large ethnic minority populations)
1987
Her Majesty’s Inspectorate of Constabulary (HMIC) produced a first
Matrix of Police Performance Indicators.
1988
The British Crime Survey recorded an apparent decline of public confi-
dence in the police. From 1988 to 1993 there was a very substantial rise
year on year in recorded crime.
1988/89
During this time there were a series of successful, high-profile appeals
against convictions for terrorist activities. These helped to throw the
probity and competence of the police into doubt (the ‘Guildford Four’, the
‘Maguire Seven’ and the ‘Birmingham Six’).
1990
ACPO published Setting the Standards for Policing: Meeting Community
Expectations (Association of Chief Police Officers, 1990). Intensified the
focus on performance and quality. It subsequently served as a cornerstone
for the Quality of Service (QOS) intitiative. These developments took
place against the backdrop of the Conservative government’s loss of con-
fidence in the police, as injections of additional resource did not seem to
bring greater public satisfaction. The ACPO was attempting to put its own
house in order, but the Conservative government eventually took its own
steps (see 1993, below).
1992
The Europol Organization was set up by the Maastricht Treaty. Europol
is based in the Netherlands and is responsible to the European Council
of Ministers. Its task is to facilitate police co-operation across the
EU, and since its foundation its mandate has been broadened several
times.
National reforms: police 55
1993
‘In 1993 the Home Secretary, Kenneth Clarke, launched a re-structuring
of police organization and accountability intended to make policing more
“businesslike” according to standards set by central government and its
local appointees’ (Reiner, 2000, p. 209).
Group 4 (a private firm) won the contract for prisoner transportation.
Publication of the Report of the Inquiry into Police Responsibilities and
Rewards (Sheehy, 1993). This called for performance-related pay (PRP),
short-term contracts and the abolition of certain management grades.
Strong resistance from police unions considerably diluted these proposals.
First set of national police performance indicators (Home Office
Circular 17/93).
Stephen Lawrence was murdered. The police handling of this racially
motivated crime drew huge criticism for its inefficiency and insensitivity.
The Audit Commission report Helping with Enquiries: Tackling Crime
Effectively, calculated that only 40 per cent of police resources go into crime.
1994
The Police and Magistrates Courts Act changed the relationship between
the Home Office, local authorities and the police authorities. The Home
Secretary became responsible for determining ‘national key objectives’.
Police authorities became more independent of local authorities, but their
funding became subject to central cash limits. The composition of the
authorities was nine local councillors, three magistrates and five appointed
members. Police authorities were to act as though they were the purchas-
ers of police services, using monitoring against performance indicators. In
general this legislation was seen as a further weakening of the local author-
ity role in policing and a strengthening of the role of the Home Secretary
(Loveday, 2000; Reiner, 2000).
Promulgation of the first set of national police objectives (SI No.
2678/1994). The national priorities were altered in 1998, 1999 and 2000
(Collier, 2006).
1996
‘The Police Act 1966 contains the seeds of conflict in so far as the Act, in
a strict sense, gives the police authority a responsibility to set priorities,
when it is the chief constable who controls resources to pursue those pri-
orities’ (Butler, 2000, p. 307).
1997
The Police Act 1997 set up the National Criminal Intelligence Service and
the National Crime Squad – important steps in the growth of a range of
56 Continuity and change in public policy and management
1998
‘Ministerial priorities’ document published. It announced that Best Value
(BV) regime would be introduced from April 2000. Under BV police
authorities had a statutory duty to consult the public in determining police
priorities:
1999
Report of the Macpherson Inquiry into the murder of Stephen Lawrence
argued that the police had been ‘institutionally racist’.
2001
Home Secretary Blunkett created a Home Office Police Standards Unit
which had an operational remit to increase police effectiveness and iden-
tify issues that might require new legislation.
2002
The Police Reform Act required the Secretary of State to produce a
National Policing Plan (NPP) each year, which would include strategic
priorities, objectives and performance indicators for police forces over the
next three years. During the first three NPPs there were quite important
changes in priorities, for example the introduction, in the 2005–08 plan,
of an objective to ‘reduce people’s concerns about crime, and anti-social
behaviour and disorder’. The Act also ushered in a new category of staff,
National reforms: police 57
2004
The new set of Statutory Performance Indicators (SPIs) signalled a shift
from a majority of process indicators to a larger share of output (for
example detections) and outcomes (for example surveys of public satisfac-
tion) indicators (Collier, 2006).
2005
Of the set of national performance indicators for 2005–08 Collier (2006)
commented:
The national targets for these priorities, particularly those derived from Public
Service Agreements (PSAs) have been politically generated rather than evi-
dence-based. The national policing plans, which began with targets for specific
volume crimes have shifted emphasis to a 15% across the board target.
● Figures for ‘total crime’ are a kind of vast aggregate, including brutal
murders and minor misdemeanours: it is often important to know
what is happening to particular parts of the picture, and modest
changes in the grand total may be fairly meaningless. (For example,
while total crime falls, and, within that, total violent crime also falls,
gun and knife crime may nevertheless go up quite sharply, leading to
conflicting headlines and political claims and counter-claims.)
● Categories of crime change over time (new crimes are defined and
legislated; sometimes old ones disappear – for example homosexual-
ity between consenting adults in many Western countries).
● Categories of crime differ considerably from one country and
legal system to another (even, for example, ‘homicide’ is classified
differently).
● The rules for counting crimes change over time (for example there
were substantial changes in England and Wales in 1998 and again in
2002 – see Hough, 2008).
● Even different police forces or units within a single country may
have different practices for recording crimes.
● Many or most crimes recorded by the police are reported by the
public. But the rate at which the public report particular types
of crime are known to vary from time to time and place to place
depending on a host of sociological and economic factors (for
example domestic violence is notoriously under-reported; theft
where an insurance claim depends on the loss being reported to
the police is likely to be well reported, at least among those social
groups who have insurance). Therefore recorded crime is unlikely to
be a very good indicator of ‘real’ or underlying crime.
● Instead of relying on crimes recorded by the police one can look
at data reported by the public – such as that in the British Crime
Survey. This should have the advantage that it includes crimes expe-
rienced by citizens which are not reported to, or recorded by, the
police. Even here, however, some will choose not to take part in the
survey, or to take part but conceal certain things that have happened
to them (Hough, 2008).
● A good deal of crime in any case follows economic cycles and chang-
ing lifestyles, and these trends cannot usually be smoothed out by
any police action. For example, economic hard times tend to gener-
ate more thefts; a society awash with small, portable phones is also
likely to suffer a range of phone-related crimes.
● In the case of police numbers, one may again find different counting
rules in different forces (Home Office, 2001).
● ‘The count of police numbers is already skewed by the changing use
National reforms: police 59
Having made all these caveats, we will nevertheless present some figures
– both in this chapter and later in the book. Readers are therefore asked
to regard them with all the caution that the above warnings suggest. In
practical terms we cannot spell out every limitation to every figure, even in
these relatively limited and selective tables – both because we cannot know
them all ourselves, and because if we did, half our book would disappear
under technical detail, which is not what we are trying to achieve.
Notes: There are minor differences between the years concerning the exact date at which
these figures were collected. Also, since the 2002 Police Reform Act certain civilian police
staff can be made ‘Designated Officers’ and allocated duties such as investigation and
detention, but they are not included in the above table.
It is also interesting that when the Serious Organised Crime Agency (SOCA) was created
in 2006 the police officers who joined it had to resign as police officers (and thereby
disappeared from police strength numbers). They became agents of SOCA, along with
ex-Customs and immigration officers and others who joined the new agency.
Both tables appear to show increases, Table 4.1 in the numbers of police
officers and Table 4.2 a far bigger increase in recorded crime. But both
tables – especially Table 4.2 – are subject to the kinds of interpretive cau-
tions mentioned earlier. The figures on recorded crime have long been the
object of criticism and interpretation from criminologists and sociologists
(Reiner, 2000, pp. 199–202). Hough makes the basic point that:
While they are often used as an index of crime, they are better thought of as a
measure of police workload – the crimes that are brought to police attention.
60 Continuity and change in public policy and management
Notes:
1. Homicides include murders, manslaughters and infanticides.
2. The statistical series ends in 2001/02 and is then replaced by a different series.
They can mislead as an index of crime because the proportion of crime reported
to the police, and recorded by them, can change over place and time. (Hough,
2008, p. 65)
Yet despite these important caveats, the trends are so broad that the
general message of Tables 4.1 and 4.2 may still be plausible: that during the
period which we cover in this book there was a huge increase in recorded
crime, and that increase dwarfed any increase in police numbers.
Certain longer-term themes are clearly apparent in the foregoing
English police timeline. One is the trend to greater central government
regulation and standardization of the police, through adjustments to the
constitutions of police authorities, through increasingly elaborate plan-
ning systems, through sharper and sharper sets of targets and performance
indicators, and through the work of the Inspectorate of Constabulary.
This trend appears to shift fairly sharply upwards from the early 1990s –
what had previously perhaps been a Tortoise becomes something rather
faster. A second is the recurrent prominence of demands for the police
to connect more closely with the concerns of local communities. This
almost seems to go in cycles, peaking every time there are serious episodes
of public disorder in specific localities. A third is the increasing profes-
sionalization and specialization of the police. This was manifested in a
myriad ways, not least the improvements in handling suspects following
PACE, but also locally (as we shall see in Chapter 4) by the creation of
special units and teams for different types of crime (gun crime, computer
crime, domestic violence, rape and so on) and for much more proactive
community relations. Taken together these innovations could be seen as a
Stalactite-like process of change.
National reforms: police 61
Since the end of the First World War there had been three main police
forces in Belgium: the Rijkswacht/Gendarmerie (national, hierarchical,
somewhat militaristic), the Judicial Police (attached to public prosecution
offices and organized in judicial districts) and the Municipal Police (under
local mayors, and with a somewhat ‘softer’ image than the Rijkswacht).
In addition there were various specialized police forces – for airports,
harbours, railways and so on – but these will not feature in our analysis.
In 1957 an Act redefined the Rijkswacht/Gendarmerie, emphasizing their
military character and replacing the 160-year-old legislation which had
stood since the period before Belgian independence. The Rijkswacht was
responsible to three ministers: Defence, Justice and Interior.
1960s
The Rijkswacht expanded and developed a Central Bureau for Investigation
(CBO – Centraal Bureau voor Opsporing) and a Surveillance and Detec-
tion Branch (BOB – Bewakings-en-Opsporingsbrigades). Technological
changes got under way, including a first radio network, the first computers
and dogs trained to seek out drugs.
1970s
Expansion and specialization. ‘More authority, more logistics, more
policemen, the establishment of specialized units, automation, centralisa-
tion’ (Bergmans, 2005, p. 1). Considerable investment was made in the
SIE/DYANE (Special Intervention Squad) which was designed to deal
with hostage-taking and other serious, organized crime. The emphasis
was on the policeman as a ‘crimefighter’. In parallel, however, a ‘turf
war’ for criminal information and the ‘best’ cases was going on between
the Rijkswacht and the Judicial Police. There were also extensive amal-
gamations of Municipal Police forces, pursuant to local government
62 Continuity and change in public policy and management
1982
There was a strike by the Municipal Police. This caused a certain resent-
ment on the part of the Rijkswacht (who were prohibited by law from
striking). It therefore did not help the attempts to bring the Rijkswacht
and the local police closer together.
1982–85
The Brabant killers (the Nivelles Gang): robbers carried out 22 armed
attacks on restaurants, stores and so on, resulting in 28 deaths and 20
injured. The gang leaders were never identified. The degree of precision in
the organization of the attacks led some commentators to claim that disaf-
fected members of the Gendarmerie/Rijkswacht were involved.
1984–85
The affair of the Cellules Communistes Combattantes (CCC): 14 bomb-
ings took place, targeted at ‘enemies of communism’. Two firemen were
killed and 28 people were injured. There was public anger when it emerged
that the terrorists had warned the authorities but the messages had not
been passed on quickly enough to the firemen. Furthermore the popular
speculation was that the terrorists must be receiving help from the national
security organization. The latter had infiltrated the cells but it failed to
take preventative action. Carette, one of the leaders of the CCC, was con-
victed for murder in 1986.
1985
Football hooligans fought in the Heysel Stadium, Brussels, during the
European Cup Final between Liverpool and Juventus. A wall collapsed,
resulting in 39 deaths and 600 injured. There was widespread international
criticism of the Belgian authorities, both for the condition of the Heysel
Stadium and for the ineffective policing of the crowd. The events of the
Nivelles Gang, the CCC and the Heysel disaster generated a good deal of
popular and political discontent with the police, and a desire for change.
National reforms: police 63
1987
The December general election returned the socialists to the governing
coalition. Louis Tobback, an entrepreneurial policymaker who wanted
change, became Minister of the Interior, and De Witte joined his cabinet
from a think tank. These two were going to be very significant players in
the future development of police policy. In opposition the Socialists had
prepared a policy perspective on the police (see Tobback, 1989).
1988
On 24 May a Parliamentary Committee of Inquiry into the Nivelles Gang
was finally set up (known as the Banditism Commission). In the same year
a major audit of the police service for the Minister of the Interior was pub-
lished (Team Consult, 1988).
1990
June: the Banditism Commission came forward with the Pinksterplan (in
English, Pentecost Plan), proposing a new emphasis on ‘community polic-
ing’ and the demilitarization of the Rijkswacht/Gendamerie. The plan was
‘clearly not a coherent master plan, but rather a somewhat unfocused col-
lection of intentions’ (Maesschalck, 2002, p. 180; Van Outrive et al., 1992,
pp. 320–21). Much of it was not implemented, but it helped set the agenda
for the later reforms (see 1998, below).
1991
July: a new law created ‘Committee P’, a standing committee of judges
and persons with police experience with an overall responsibility for moni-
toring Belgium’s police forces and intelligence services. Ever since then
Committee P has issued detailed reports coverning many aspects of polic-
ing (see www.comitep.be). General elections saw the rapid rise of far right
parties. The first law relating to police reorganization was proclaimed,
coming into effect in 1992.
1992
Second ‘Pinksterplan’: Municipal Police forces were to be required to
cooperate with each other in ‘interpolice zones’. Instead of 589 Municipal
Police forces there would be fewer than 200 such zones. Also introduced
was a new crime prevention policy based on ‘local security contracts’.
1992 Implementation of the Police Function Act, which re-defined
the respective functions of the municipal police, the Rijkswacht and the
64 Continuity and change in public policy and management
judicial police. This was intended, inter alia, to promote greater integra-
tion between the municipal police and the Rijkswacht.
1995
The first (voluntary) initiatives with respect to ‘interpolice zones’ –
attempts to improve coordination between forces at the local level. The
provincial level played an active role in stimulating local collaboration.
De Witte (see above) became Governor of Flemish Brabant, and was an
influential figure in the police policy process.
1996–97
The Dutroux Affair hits the headlines. Marc Dutroux was a paedophile
who sexually abused six girls, killing four of them. The case showed gross
inefficiency and a lack of coordination – indeed, sometimes active hostility
– between the judiciary and the three police forces and among the police
forces themselves. There were also suspicions of police corruption and
complicity. The Dutroux Affair triggered the ‘White March’, the largest
public demonstration in Belgium since the Second World War, when
250 000 to 300 000 people marched through Brussels to express discontent
with the handling of the affair by the police and the courts. This came as a
major shock to the public authorities, and finally led to substantive change.
A Dutroux Parliamentary Committee of Inquiry was set up almost imme-
diately, and proved to be the final catalyst for real police reform.
In April 1997 new legislation gave mayors a say in local branches of the
(national) Rijkswacht/Gendamerie.
1998
In April Dutroux briefly escaped from custody. The Ministers of Justice
and of the Interior immediately resigned. Prime Minister Dehaene invited
the four opposition parties to join the four government parties to discuss
reforms in the criminal justice and police system. These eight-party talks
led to the ‘Octopus Agreement’ of 24 May. This proposed an integrated
police force on both national and local levels.
1998
The Police Reform Act (which resulted directly from the ‘Octopus
Agreement’) grouped together three previous polices forces (Municipal,
Judicial and Gendamerie). The Judicial Police had not originally been
part of the main story, but: ‘There was a conflict between the BOB [special
investigation branch of the Rijkswacht] and the judicial police. Tobback
and Michiels agreed that something had to happen also with the judicial
police’ (interview 32 – Michiels was a senior and influential police officer).
National reforms: police 65
1999
General elections saw the arrival in power of a ‘rainbow coalition’ of liber-
als, socialists and environmentalists.
2000
May: Federal Security Plan and Prison Policy.
December: two ex-Gendarmes/Rijkswacht appointed, as General
Superintendent of the Federal Police, and General Inspector of the
Federal and Local Police.
Euro 2000 Football finals were held in Belgium and NL amidst inter-
national security concerns. Coincidentally, these fears strengthened the
hands of the police unions, who threatened to strike during the cham-
pionships and eventually secured generous pay deals for the new police
structure.
2001
Beginning of implementation of 1998 Police Reform Act. Also a royal
decree organizing police training establishments. ‘[T]here is today an
unambiguous move towards organizing the police along “managerial”
lines’ (Tange, 2004, p. 233).
2002
Establishment of 196 local zones for the police (see entry for the 1970s,
above).
2004
A ministerial circular announced the introduction of organizational devel-
opment and quality management programmes for the police. The intro-
duction of an integrated security framework also meant that, for the first
time, combating crime was seen as a task involving other partners beyond
the police and the criminal justice system.
2005
The Vesalius Acts sought further to integrate existing police personnel into
the new structure of local and national forces. These harmonization meas-
ures turned out to be rather expensive. At this point there were roughly
13 000 federal police and rather more than 30 000 local police. In addition
66 Continuity and change in public policy and management
there was a different debate between Flanders and Wallonia. However, there
was not a problem with the Gendarmerie, which was dominated by Flemish
generals (Berckmans, De Ridder, Franssen) which received management train-
ing at the Vlerick School. (interview 32)
Table 4.4 Numbers of police officers per 100 000 population, 2003
Source: European Sourcebook of Crime and Criminal Justice Statistics (2006), p. 74.
Even if the English story is more hectic, more crowded with new policies
and initiatives, there are some commonalities between the two countries.
In both, failures to solve headline crimes (or failures to solve them fast
enough) brought demands for reorganization. In both there were instances
not just of incompetence but also of corruption, or presumed corrup-
tion, which took their toll on public attitudes towards the police. In both
there were periods of difficulty when the police were brought in to deal
with large-scale public order problems (miners’ strikes, and so on) which
took on an overtly political character. In both there was a steady ten-
dency towards larger geographical units for police operations, although,
as explained earlier, the main English forces are much bigger than the
municipal forces in Belgium, but lack the national reach of the Belgian
Rijkswacht/Gendarmerie. The interest in having bigger units was pro-
pelled by several common motives, including a belief in economies of scale,
better geographical coordination, and improved management and techni-
cal services. This essentially structural approach was probably a more
prominent part of the Belgian reform process than the English. Indeed, as
we shall see in our local study (Chapter 7) an enormous amount of effort
went into renegotiating local Belgian police boundaries and coordination
National reforms: police 69
of coloured jacket with their role labelled on it. There are, therefore, some
(rough) parallels between England and Belgium, even if the details and
degree of differentiation within law-and-order work are substantially dif-
ferent in the two countries.
Nevertheless the sheer pace of change in England, especially since the
late 1980s, demands attention. The managerialist apparatus of plans,
standards, targets and performance-linked budgets – although present to
some extent in Belgium – arrived sooner and has been developed much
further in England. The application of high-tech, specialized techniques
of information management, force deployment, crime scene analysis
and neighbourhood policing all appear to have flourished in England
beyond anything we encountered in Belgium. On the other hand, it could
be argued that much of this managerial apparatus was designed at least
partly to improve the ability of UK ministers to see and control what the
police were doing, while in Belgium political control was already present,
though in a quite different form. Certainly the influence of local mayors
over Municipal Police forces has no parallel in England. In the latter, it
is true that local councillors sit on police authorities, but this is a far cry
from the kind of detailed, single-minded influence that can be achieved by
a determined Belgian Mayor.
5. National reforms: intersectoral
comparison
5.1 SIMILARITIES AND DIFFERENCES IN BRITISH
AND BELGIAN POLICYMAKING
To some extent, but only to some extent, the opening stereotypes (Chapter
2) do appear to have survived closer scrutiny. Thus there has indeed been
a significant contrast in policymaking styles and intensity between the two
countries – in respect of both police policy and hospitals policy. Given that
the two countries faced many similar problems, this degree of difference
stands as a remarkable tribute to the divergent influences of the respective
cultural and political frameworks. England has been the site of relentless
activism, with one reform following another. Belgium has been much
slower-moving, with big reforms emerging only occasionally, usually after
long periods of public indignation and political pressure. Graphically, one
might portray the different frequencies and amplitudes as the kind of ‘car-
diogram’ beloved of TV hospital dramas – see Figure 5.1
Belgian
police
English
police
Belgian
hospitals
English
hospitals
71
72 Continuity and change in public policy and management
What kind of changes have these been? Here we may first return to
the three-level scheme of Hall introduced in Chapter 1, section 1.5. The
first (most detailed) level was changes in the levels set for specific policy
instruments. Here there have been many changes in both countries – in
fact these have been far too numerous to reiterate them all here. For the
English police the 1979 pay increase was significant, and the more recent
annual adjustments of performance targets also falls into this category.
Meanwhile in Belgium there was the gradual demilitarization of the
federal police and the equally gradual professionalization of the local
police (more training, more specialization). In English hospitals the period
since 1997 has undoubtedly witnessed a large, policy-driven reduction in
average waiting times for elective inpatient treatment. In Belgian hospitals
the 1982 moratorium on opening new hospitals beds or the 1993 increase
in patient co-payments for visits to doctors represented changes in the
level at which policy instruments were set. Overall, the greater activism in
England has meant more specific changes of this kind, but neither country
has been at all static.
On the second level Hall (1993) was concerned with shifts in the type of
policy instrument used. This is a bigger type of change and therefore it is
not surprising to find fewer shifts of this kind. For the English police we
could cite the multiplication of specialist bodies at national level, includ-
ing the 1997 National Criminal Intelligence Service and the 2004 Policing
Improvement Agency. For the Belgian police there was the introduction
of local security contracts from the early 1990s. In the English hospital
service the increased use of market-type mechanisms has been a notable
innovation – first following the Conservative government’s reforms of
1989 (the ‘provider market’) and then more recently under the banner
of the New Labour government’s commitment to ‘choice’. Among the
Belgian hospitals – where ‘choice’ had always been part of the system – one
could say that the idea that the whole population of hospitals constituted
a system which required federal-level planning was a change which gradu-
ally grew during the 1970s and 1980s, and required the federal government
to fashion new regulatory instruments. On this level the contrast between
England and Belgium appears rather clear. The second-level changes in
England were more marked and more numerous than those in Belgium.
Hall’s third level denotes paradigmatic change – a new concept of
the policy and/or its objects. Such changes are not common – at least
not with established services such as the police and hospitals. They may
occur incrementally over time (Stalactite-type changes in Table 1.1) or as
an Earthquake (punctuation). It is hard to see such changes in Belgium
at all. Perhaps one candidate would be the role of management itself:
in the Belgian hospital sector hospitals gradually became entities which
National reforms: intersectoral comparison 73
public or the policymaking elite. Second, in the English health care arena,
there has arguably been a fundamental shift away from ‘doctor power’,
almost to the point where the medical profession is seen as the problem
holding back the development of a more efficient and effective health
service. The demands for better management and more clinical transpar-
ency have been endless, and although the medical profession has been a
powerful defender of its autonomies, since around 1990 it has been obliged
to cede a great deal of ground. Like a Stalactite, management has just
gone on growing. Again, while some of the arguments can also be found
in Belgium, we see no equivalent shift – the talk has not yet become the
practice or the policy.
We must be careful, however, not to read strong normative lessons into
these contrasting histories. We should not assume that English hospitals
are therefore better than their Belgian counterparts, or that the English
public are happier with their hospital system. Neither should we assume
that crime is better tackled in England, or that the public fear of crime is
lower there, or even that the many English community policing initiatives
and programmes, however enthusiastically pursued, have ensured higher
levels of public order and public trust in the police. Indeed, we will now
look at some measures which seem to show a rather more complicated
picture.
Before we come to the figures themselves we need to acknowledge their
limitations, which are many. Sometimes the relevant figures are available
only for the UK, or for England and Wales, rather than for England
alone. Sometimes the figures are available only for one year – when a par-
ticular study was undertaken – or are available for only a small subset of
our 40-year timespan from 1965 to 2005. Often categories do not fit both
countries in quite the same way (for example with respect to questions
about the affordability of healthcare, where the payment systems in the
two countries have been so very different). Frequently, also, when surveys
tap public perceptions, the responses may be conditioned by cultural
expectations and norms specific to that country. For all these reasons,
therefore, we must be careful not to put too much weight on comparative
statistics. On the other hand, it would seem foolish to ignore them. So we
have chosen what seemed to us the most relevant and interesting compari-
sons, and treated them as much as a source of further questions as a set of
definitive answers.
First, then, let us look at certain aspects of healthcare and hospitals (see
Table 5.1).
The Belgians, it seems, spend rather more on their health care system.
It appears to be a more generous, less efficient system (in the sense of
having more nurses, doctors and hospital beds per capita) than that in the
National reforms: intersectoral comparison 75
Aspect Belgium UK
Total expenditure on health as a % of GDP, 2002 9.1 7.7
Total public expenditure on health as a % of GDP 6.5 6.4
Total private expenditure on health as a % of GDP 2.6 1.3
Acute hospital beds per 100 000 population 583 390
Average length of stay (days) 19961 7.5 4.8
Nurses per 100 000 population 1075 497
Physicians per 100 000 population 448 390
Life expectancy 78.4 78.2
Infant deaths and neo-natal deaths per 1000 live births 9.2 8.7
% of people who perceive their own health as good or 78.2 74.0
very good
Source: World Health Organization Europe (2006), pp. 10, 11, 31, 33, 47.
UK. It is true that long English hospital waiting lists, which used to be
characteristic of the UK (but not of Belgium) have been much reduced by
the intense reforms (and large expenditure increases) of the New Labour
government after 1999. A Eurobarometer survey in 2007 found that 87
per cent of Belgians found access to their hospitals ‘very easy’ or ‘fairly
easy’ but that the UK was not that far behind, with 80 per cent. Against
that there is the financial aspect. NHS hospitals are free at the point of
access (financed principally from general taxation), whereas in Belgium
insurance is normally required. The same survey found that 31 per cent of
Belgians thought healthcare was ‘not very affordable’, or ‘not at all afford-
able’, while only 6 per cent of UK citizens fell into this category (Special
Eurobarometer, 2007, pp. 28–32).
Health status outcomes are not strikingly different between the two
countries, but Belgians do feel a bit healthier than the average UK citizen.
On the other hand we need to treat life expectancy statistics with the
knowledge that probably only about one-quarter of the health of people
in developed countries can be attributed to the doings of their health-
care systems. The rest can be traced to the physical, economic and social
environments (which influence lifestyle choices), genetic inheritance and
human biology (Canadian Institute for Advanced Research, 1997).
Now we turn to the police. In Tables 5.2 and 5.3 we see further evidence
that policy activism does not automatically translate into better outcomes.
In Table 5.2 we see that there are many more Belgian police per capita than
English and Welsh police per capita. It also appears that ‘civilianization’
76 Continuity and change in public policy and management
Note/Source:
1. Average per year, 1999–2001. The EU average was 337. Source: Barclay and Tavares
(2003).
Measure Belgium UK
Victimization rate for 10 common crimes in 2004 6 2
Car theft in 2004 12 1
Burglary victimization rate 7 1
Assault and threat victimization rate 4 1
Contact with drug problems in area of residence 9 6
Feeling unsafe on the street after dark 12 8
Victims satisfied with police after reporting crimes 11 9
(lower number means more dissatisfaction)
Note: The rank of 1 = highest crime in the EU-18 and a rank of 18 = the lowest. Thus,
for example, the UK has the highest reported rate of car theft among the 18 EU countries
surveyed.
has gone much further in England and Wales, where in 1995 there were
106 civilians working for the police per 100 000 population, compared
with only 40 in Belgium. The EUICS Consortium report The Burden of
Crime in the EU (2005) shows that, in general, crime rates have been falling
throughout most of the EU since the end of the 1980s. Belgium is unusual
as the only country where crime has actually slightly increased during that
period. However, before any assumption is made that this means that the
serially reformed UK police are doing much better than in Belgium, look
more closely at Table 5.3. Both Belgium and the UK are fairly high-crime
countries by overall EU standards – they are both in the top halves of most
of the tables in the report. But within that unfortunate category, Belgium
comes out better on all four of the common crime rates shown in the table
than the UK. Furthermore, the people surveyed in the UK had more
contact with drug problems where they lived, and greater fear when out in
National reforms: intersectoral comparison 77
the street after dark. They were, however, more satisfied than the Belgians
with what happened when they reported crimes to the police. That may
in part be because, when victims of crime, they received somewhat more
support – see Table 5.4. However, all these crime figures should be read
with all the caveats already discussed in Chapter 4.
When we come to the issue of general public trust in the police (Table
5.5), we find a large contrast. Despite the various scandals and problems
alluded to above, the majority of the British public evidently trust their
police. In Belgium, however, confidence in the police was for some years
at one of the lowest levels in Western Europe, although it has recovered
somewhat, and by 2005 was only slightly below the EU average (whereas
in the UK confidence in the police is five percentage points above the EU
average). In both countries confidence in the police is considerably higher
than confidence in the legal system.
In short, it would be a gross oversimplification to assume that greater
policy activism usually leads to either better policy outcomes or more public
satisfaction. There are far too many other factors that may influence these
outcomes, including, obviously, how sound the policies are, how well they
are implemented, what else is happening in the social and political envi-
ronment at the time, and what the relevant public expectations and social
norms may be. (For example, the exceedingly low confidence score of the
Belgian police in 1997 – see Table 5.5 – was almost certainly influenced
by the fact that the Dutroux affair was headlines at that time.) Crime is
influenced by many factors largely beyond police control (such as the eco-
nomic cycle and lifestyle changes) and health outcomes by many influences
beyond the control of hospitals (culturally influenced patterns of diet and
exercise, levels of pollution). Furthermore, each country starts from a dif-
ferent place – a different current norm. Thus crime may have fallen faster
in England and Wales than in Belgium during the period 1988–2004, but
that still left absolute crime levels higher in England and Wales, even after
a decade and a half.
78 Continuity and change in public policy and management
Note: The percentages in the table are those who said they tended to have confidence in
the police and the legal system.
Final
outcomes
Socio-economic (inpacts)
Needs
problems
Intermediate
outcomes
(results)
Organization or programme
Objectives Inputs Activities Outputs
Relevance
Efficiency
Effectiveness
Utility and sustainability
least two counters to such an argument. First, it could be that the more
‘relaxed’, better-resourced nature of the Belgian hospital system reflects,
however imperfectly, a collective Belgian preference that this should be
so. That Belgians do not want their hospitals to be straining for the lowest
possible length of stay, sending patients home as soon as is possible so that
beds can be used with maximum efficiency. That they want the choice that
comes from a looser system (where they do not necessarily even have to
go through a primary care doctor to get to hospital). Perhaps they are pre-
pared to pay for more slack in the system, and/or they are prepared to vote
for politicians who will act to preserve that higher level of provision. (Or,
at least, to vote against anyone who seems to want to cut back on hospital
spending.) Second, there is, of course, no guarantee that resources released
from efficiency gains in the hospital sector will be used for noble public
purposes. They might equally be used to grant additional tax breaks that
wealthy people seize upon to buy bigger motor cars and drink more cham-
pagne, or indeed to finance public sector programmes which voters may
not approve of, such as lavish office blocks for civil servants or motorways
through environmentally sensitive localities.
These are fairly high-level arguments (although none the less important
for that). In addition, there are practical complications. Looking at the big
picture (Table 5.1) it may appear that UK hospitals must be more efficient
than their Belgian counterparts. But how and why? Unless one has much
more specific information, it will be very difficult to fashion pro-efficiency
80 Continuity and change in public policy and management
81
82 Continuity and change in public policy and management
many again. Along with Brugge, Gent and Mechelen it is one of Belgium’s
ancient Flemish cities, full of beautiful old buildings, including a spectacu-
lar fifteenth-century Gothic town hall. It is also a university town, housing
Katholieke Universiteit Leuven (KUL), founded in 1425 and the oldest
university in the low countries. The 35 000 students give a youthful air to
the town, and a cosmopolitan one too, since more than 12 per cent of them
currently come from outside Belgium. Leuven is also influenced by being
only a 20-minute drive (17 kilometres) to the capital, Brussels.
In this chapter we will tell the stories of 40 years of development in the
Brighton and Leuven hospital systems. These stories will be compared,
one with the other, and particular attention will be paid to how far devel-
opments in these two localities can be ‘read off’ from the national policies
discussed in Chapters 2 and 3, and how far they seem to have had a more
distinctively local character.
For most of the period since the mid-1960s Brighton has had two medium-
sized general hospitals, the Royal Sussex County Hospital (hereafter
RSCH) and the Brighton General. Both are nineteenth-century founda-
tions, both have rather untidy campuses. Neither has ever been fully mod-
ernized. Since the mid-1960s the relevant authorities have made two or
three attempts to get a new hospital built on a new site, but each attempt
has failed. As the Chair of the Brighton Health Authority put it in her 1985
New Year message to staff, local acute health care was constrained by ‘the
enormous obstacles presented by our obsolete buildings. A new hospital is
essential’ (Cumberlege, 1985). Six years later Brighton’s medical consult-
ant staff wrote a joint letter to the then Conservative Secretary of State for
Health saying:
To our dismay we . . . hear that any money set aside . . . for Hospital Building pro-
grammes in Brighton has now disappeared. New building worth £51 million had
been approved in principle five years ago. Our three remaining acute hospitals:
But the new hospital did not come. Instead, a number of small hospitals
have been closed and acute services have been incrementally concentrated
What happened locally? Hospitals 83
on the increasingly crowded RSCH site. Since the late 1990s a good deal of
new investment has been put into that site, but it was still the case that, at
the end of 2006, 209 beds were in buildings constructed before 1850, com-
pared with only 277 in those built since 1960. Meanwhile acute services at
the Brighton General have been run down, and no advanced surgery has
been conducted there since the early 2000s.
This has been, therefore, a local history of slow, Tortoise-like incre-
mental change, false starts and disappointments, but a definite underly-
ing trend towards concentrating ‘hot’ services at the RSCH. In 2005 the
RSCH suffered the ignominy of simultaneously failing to gain any stars
at all in the government’s hospital performance league tables (The Argus
2005, p. 1), developing one of the largest budgetary overspends of any
hospital in the country (Brighton and Sussex University Hospitals NHS
Trust, 2005), and being the subject of an ‘undercover’ BBC TV documen-
tary showing nurses ill-treating elderly patients on one of its wards (BBC
News, 2005). Significant bed closures followed as management attempted
to right the financial position. Thus, by the end of our period of study
(1965–2005) Brighton’s hospital services could hardly be termed a ‘flag-
ship’ for the NHS.
We now turn to Leuven. A major turning point here came right at
the beginning of our period – the events of 1968, known in Flemish
as de splitsing – the splitting-up. What was splitting was the university
(Katholieke Universiteit Leuven, or KUL), as the Francophone academ-
ics and students were virtually expelled, and left to set up an alternative,
Francophone (Walloon) university 50 kilometres away at Louvain-la-
Neuve. The Flemings remained in situ, and that applied as much to the
Leuven hospitals and the medical school as to the rest of the university.
The whole drama was part of the larger conflict between the two major
language groups in Belgium – the Flemings and the Walloons.
Shortly after de splitsing Piet De Somer, the then Rector of KUL,
appointed an enterprising young doctor, Jan Peers, as Director of Medicine
and General Director of the St Rafaël Clinic. Building on previous plan-
ning and extensive investigations of alternative hospitals systems by his
predecessor, Jan Blanplain, Peers soon became the formative influence
in pushing for a new, single-site hospital on a low hill just outside the old
city, at Gasthuisberg. Within a decade a huge new hospital, designed on a
low-rise, key pattern was beginning to manifest itself on the site. This was
a startling new development, a real Earthquake. Formally inaugurated
on 26 January 1985, the new UZ Leuven (University Hospital Leuven) is
now the biggest hospital in Belgium (1800 beds by 2003) and in a recent
survey of primary care doctors’ opinions scored as the best teaching hos-
pital in Belgium in 12 out of 13 medical specialties (Test Gezondheid,
84 Continuity and change in public policy and management
It should be noted that this history is being given a very particular focus
– one concerned with the development and decay of physical and tech-
nological facilities, and of organizational forms. New hospital buildings,
technological advances, changes in organizational structures and financial
constraints are the prime concerns. These are, of course, far from being
the only aspects of hospital life. A great deal depends on the skills and
management of staff, the availability of finance, changes in the patterns of
disease and hospital usage, and so on. These factors will not be ignored,
but they will be treated mainly in relation to the prime focus.
However, a firm caveat needs to be entered to the effect that good medi-
cine and nursing can be (and frequently are) practised in old buildings, and
even with poor local management. Equally, local managers may be doing
their remarkable best yet be undone by ill-considered national reforms
which are forced upon them, or by low-quality clinical practice within
their institutions (which managers can seldom, if ever, closely control).
The latter may never be discovered, or (increasingly likely these days) it
may bring both political and media criticism down upon their hospitals.
Nothing written here should be interpreted as a judgement of particular
individuals.
Sources are a crucial feature of any historical account. In this case we had
generous access to both persons and papers, although the nature of the
documentation differed somewhat on either side of the Channel.
One reason for choosing the period since the 1960s was that a good
proportion of the key decision-makers were still alive and potentially
What happened locally? Hospitals 85
available for interview. In the event, we were able to interview all the
hospital chief executives for both the Brighton and Leuven hospitals for
virtually the whole period, plus a good number of other senior figures,
including some with broader political responsibilities. In all we conducted
17 interviews with the key players and observers, using a standard sched-
ule of questions but departing from that if the respondent wanted to lead
us onto new or different issues. Records of these interviews were usually
sent to the respondents so that they could correct any mistakes and add
further thoughts if they so wished. We also examined a large number of
documents. In Leuven we were able to see speeches and policy papers,
consultancy reports, an example of a strategic plan and a number of retro-
spective accounts produced for the 75th anniversary of the foundation of
the university hospital in 2004. In Brighton the documentation was more
extensive. Hospital Board minutes and planning documents were available
back to 1993. A full set of the monthly local Health Bulletin was analysed
back to that journal’s foundation in 1967. The Brighton newspaper, The
Argus, has an archive which enabled us to track down reports on the hos-
pital going back to 1985. The far greater public documentation for the
Brighton case probably reflects the fact that it is a unit within a publicly
accountable National Health Service, whereas Leuven University Hospital
is a non-profit foundation, subject to government regulation but not a
direct part of the state apparatus, even if it treats and relies upon public
patients largely paid for through the Belgian national health insurance
system. But it also reflects a more general difference in political cultures:
as one of our Leuven interviewees put it to us in mid-2006: ‘the debate on
public disclosure is only just starting in this country’ (interview 9).
The research was carried out from 2006 to 2008.
at HM Treasury. One of the main reasons for this was that one thing new
hospitals inescapably require is capital – and lots of it – and in the NHS
system capital could only be obtained from above. However, the authori-
ties ‘above’ had far more to worry about than Brighton’s wishes and
needs. They had to look at which were the most urgent needs across the
whole region. The ministry had to look at what were the national policy
priorities. The Treasury had to look at the state of the economy and of
public finances, and to take consequential decisions about what should be
the overall level of public investment. Multilevel, networked governance is
nothing new to the NHS.
So why didn’t the ‘higher levels’ furnish the capital Brighton so clearly
needed and demanded? It appears that there is at least a three-part answer
to this. First, Brighton may have been needy, but it was not the neediest
place in the area or region. Two other south coast towns – Eastbourne
and Hastings – had hospital facilities that were judged to be in even worse
condition, and they received priority capital to fund major new develop-
ments (interview 3). Second, Brighton was singularly unfortunate (or
tactically clumsy) with the timing of its bids for new hospital development.
The mid-1970s bid for a new hospital on the Falmer site (out of town, next
to the university campus) coincided with the biggest public expenditure
crisis since the Second World War. (Figure 6.1 shows the locations of
the principal sites in this narrative.) Humiliatingly, the Labour govern-
ment of the day had to go to the International Monetary Fund for a loan,
and part of the price of that was a particularly savage series of public
expenditure cuts. New public sector investment virtually disappeared for
a while, the Falmer hospital included (interviews 3, 21). The second bid,
in the late 1980s, combined a major redevelopment at the RSCH with a
brand new medium-sized hospital at Holmes Avenue in Hove (the site was
never going to be big enough to have a full-scale general hospital there).
The bidding process dragged on for years, and in the end it too ran into
a period of strict national capital rationing by the Treasury (accompa-
nying the severe economic downturn at the end of the 1980s). The third
reason was that the local elites never spoke with one voice – there was
constant internal controversy. As one ex-Chief Executive said, despite a
careful analysis of many possible sites, ‘the harsh reality was that nobody
could agree’ (interview 18). To put it crudely (the actual dispositions were
somewhat more subtle) local Brighton politicians were nervous about
the closure of the old Brighton General while a new hospital was built in
the rival town of Hove. Meanwhile the Hove politicians were keen on the
Hove (Holmes Avenue) development, although many of the doctors were
publicly critical of the impracticality of having two ‘hot’ sites (at RSCH
and Holmes Avenue) because of the staffing and time implications of
88 Continuity and change in public policy and management
kilometres HAYWARDS
HEATH
0 10
KEY
SUSSEX PRINCESS
BUILT-UP COUNTRYSIDE ROYAL
AREAS HOSPITAL
HOLMES AVENUE
SITE FALMER
SITE
BRIGHTON
GENERAL
HOSPITAL
HOVE
SHOREHAM
BRIGHTON
ENGLISH
CHANNEL
ROYAL SUSSEX
COUNTY HOSPITAL
from 1997 Whitehall’s grip intensified. In the late 1990s and early twenty-
first century there was ‘much more interference and even bullying’ (inter-
view 3). One experienced ex-Chief Executive acknowledged that nowadays
‘there was an obsession with top-down performance management’, and
that under Alan Milburn’s period as Secretary of State (1999–2003)
central ‘performance management went into orbit’ (interview 18). ‘We are
now performance managed to the nth degree’ (interview 18). Nationally
(as we saw in Chapter 3, Table 3.2) many chief executives had lost their
jobs for perceived performance failures, and fear and disillusionment had
become widespread. There grew up ‘a culture of bullying’ by the ministry
and their regional offices: ‘it is so counterproductive’ (interview 22).
One brighter note in the history of the Brighton hospital system was
the setting up, in 2001 of a medical school, linking the hospital to the two
Brighton universities. Again this was connected to central government,
in the sense that the latter had decided to encourage a limited number of
new medical schools, and Brighton’s bid was (eventually) one of the suc-
cessful ones. The first attempt to bid failed because parts of the University
of Sussex (seen by the government as an essential partner) were ‘as cold
as custard’ towards the idea (interview 15). Eventually, however, the aca-
demic politics of it was sorted out, and a successful application, binding
the hospital trust and the universities of Sussex and of Brighton, was made.
The existence of a medical school not only gave an academic flavour to
medical life, it also boosted research, brought in extra money, offered new
possibilities for clinicians who wanted to develop beyond their day-to-day
clinical practice and was ‘really, really good for morale’ (interview 22).
However, by 2005, as indicated earlier, the trust was in crisis, hit by a
large overspend, a score of zero stars in the government’s performance
league tables, and a deeply embarrassing undercover TV documentary
showing indefensible nursing practices on a geriatric ward. How had this
come about? One persuasive story is that the trust had entered a period of
instability, during which management had ‘taken its eye off the ball’ (inter-
view 25). Previously, there had been only three chief administrators and
managers since the NHS was created back in 1948. Then the incumbent
Chief Executive (CE) had fallen ill and there had suddenly been a succes-
sion of three acting or temporary CEs in the space of 12 months. Moreover,
during this time there had been a number of major new developments which
had distracted attention from the basic business of controlling finance and
monitoring the quality of the hospital’s services. First, there was a merger
of Brighton with Mid-Sussex, bringing the Princess Royal – quite a modern
hospital at nearby Haywards Heath – into the same organization as the
RSCH. Second, there had been the arrival of a Medical School, as described
in the previous paragraph. Initially, this had been a statutorily separate
What happened locally? Hospitals 91
Again, the main events are summarized in tabular form (Table 6.2).
A federal-level hospital policy began to emerge in Belgium during the
mid-1960s – just prior to the launching of the idea for a new, greenfield
site KUL Hospital. The first Hospital Act came at the end of 1963, and
introduced the idea of planning the hospital system as a whole. The first
national plan itself appeared in 1966 (European Observatory on Health
Care Systems, 2000).
The early years of the project for a new KUL hospital were remark-
able in many ways. De splitsing was a national political crisis, and one
where the most dramatic manifestations were centred on KUL. The
Flemish politicians and academic leaders were, as they saw it, decisively
ending more than a century of dominance of their half of the country by
a patronising Francophone elite. This provided the immediate backdrop
for the emergence of the new hospital, envisioned from the outset as an
international-class Flemish teaching hospital. ‘We can do it as well as the
French universities’ (interview 5).
As usual, political opportunity favoured the prepared. During the 1960s
the Director of St Rafaël had carried out a study of 55 university hospitals,
looking for the most suitable solution for future development (Blanplain,
What happened locally? Hospitals 93
2004, p. 9). (St Rafaël was one of two university hospitals within the old
city of Leuven, and attached to KUL – St Rafaël was the Flemish one
and St Pieter was the Francophone one. Geographically they are virtu-
ally back-to-back. This twinning symbolized the divisions between the
two language communities in Belgium.) So at the time of de splitising
the senior medical establishment at KUL already had ideas for the way
ahead. Furthermore, de splitsing led to the need for the Francophones
ejected from Leuven to redevelop their medical facilities, which they did by
building a new university at Louvain-la-Neuve and a new medical school
at Wouluwe St Pierre, an eastern suburb of Brussels. These new projects
involved considerable federal government finance and so, by the unwritten
rules of the Belgian political game, the Flemings were also entitled to ask
for something comparable (interview 9). And at that moment there were
no serious Flemish rivals – no other Flemish university was proposing a
major new tertiary hospital. KUL was, in fact, in a strong position. Not
only had it been so centrally involved in de splitsing, but it had also been
responsible for training more than two-thirds of Flemish doctors.
So the politics of the situation were crucial. Once the federal government
94 Continuity and change in public policy and management
gave its permission for the building of a new hospital, everything else fol-
lowed. The federal government provided a 60 per cent subsidy itself and
the remainder was borrowed from banks, but these loans were reimbursed
over a 30-year period through an element in the running cost payments
which the hospital received – also from central government. Getting
government permission was eased by the coherence and relatively small
size of the Flemish elite. In the allocation of ministerial portfolios at the
federal level the Ministry of Health usually went to a Christian Democrat
(at least until the late 1980s) and the new UZ at Leuven was every inch
(or we should say every millimetre) a Christian Democrat (CD) project
(though not only a CD project) (interviews 6, 10). KUL and the medical
school were closely connected not only to the political party but also to
the Roman Catholic hierarchy (the university’s Chancellor is, ex officio,
the Cardinal Archbishop of Mechelen-Brussel). All these fractions of the
Flemish elite could be expected to welcome the idea of a major new hospi-
tal run by a Christian institution rather than by the state. It was possible, at
least temporarily, to achieve a unity – or near unity – that would have been
far more difficult in a much bigger, more diverse country like the UK.
However, not everything was plain sailing. There was one rival – an
internal one – for the scheme to build a brand new hospital outside the
city at the St Gasthuisberg site. This was a proposal to rebuild and merge
the two inner city hospitals, St Rafaël and St. Pieter’s. It was supported
by a group of Flemish medical staff led by the then Dean of the Medical
faculty. The contest was settled in a way unthinkable within the hierarchi-
cal British NHS – by allowing both projects to proceed until one of them
became an obvious winner and the other an obvious loser. That is why, in
2006, it was still possible to walk around an empty, decaying 1970s tower
block which was to have been part of the new inner city hospital, had it
not become clear that: (1) there would not be enough money to finish
both projects; and (2) the Gasthuisberg project embodied more modern
and progressive concepts of care, and commanded a stronger network of
support, including the Minister and the Rector of KUL (interview 9).
From the mid-1970s to the late 1980s the new hospital grew and grew,
and its finances remained tolerably healthy. There were fluctuations, with
downswings into losses in the early 1980s and early 1990s, but this pattern
was understood to be largely the outcome of a system in which university
hospitals developed new treatments and technologies and were only fully
reimbursed for these some years later, when the federal payment authori-
ties had adjusted their payment categories to accommodate the novelties.
This decade was, therefore, something of a golden age. At the formal
inauguration in January 1985 the KUL Rector referred to the ambitious
new hospital as the result of nearly 20 years of planning, and thanked the
What happened locally? Hospitals 95
Every Western government – not only in Belgium and the UK – has been
obliged, from the late 1970s onwards, to attempt to restrain the growth of
public spending on health care. It is an international problem, endlessly
discussed internationally as well as nationally. As we observed in Chapter
1, this huge fiscal pressure is fuelled by a number of virtually universal
trends. People live longer and the rapidly developing medical technologies
can do more and more for them. As several of our interviewees confirmed,
the expectations of hospital patients have also risen. They expect prompt
What happened locally? Hospitals 99
Having set out national police policies in Chapter 4 we will now turn to
two local police jurisdictions, to see what policy change (or policy stasis)
looked like closer to the ground. As in Chapter 6, we will look at Brighton
and Leuven, although here we immediately encounter one important
difference between the two countries. Whereas Leuven has had its own
police force since 1796, Brighton is only a division within a much larger
police area – the Sussex force. This is typical of the more general national
patterns – police areas in England are much bigger than in Belgium. As
we will see, this immediately makes for substantial differences in terms
of the size of forces, the levels of internal specialization, and the range
of technologies and problems with which the ‘local’ police deal. This size
difference is partly a reflection of the fact that, whereas the local Belgian
forces had always operated with the umbrella of a national police force
(the Rijkswacht), England has never had a national force. (However, as
we saw in Chapter 4, it has developed more and more specialist units at
the national level, such as the Serious Organised Crime Agency created
in 2006 or the National Policing Improvement Agency established under
the 2006 Police and Justice Act. But these are not police forces, they are
special units.)
We should first say that we are extremely grateful for the generous coop-
eration we received from the Sussex Police and the Leuven Police (De
Leuvense Politie). They made available documents and records and also
afforded us 20 lengthy, semi-structured interviews with senior officers and
political leaders, both past and present (12 of these were in Sussex, eight in
Leuven – for further details see the Appendix).
As with hospitals, one immediately noticeable difference between the
two organizations was the amount of detail and the number of documents
in the public domain. English police forces publish more reports and more
detailed reports than do their Belgian counterparts, and many of these
101
102 Continuity and change in public policy and management
The Sussex Police area is full of contrasts. It has two urban centres with
pockets of poverty and relatively high crime rates – Brighton and Hastings
(see Figure 7.1). Brighton has the second-largest police station in England,
and the post of Divisional Commander there is coveted as a career stepping
stone to the Assistant Chief Constable level. Most of Sussex, however, is
quiet and rural – and fairly rich. But in the north lies Gatwick – London’s
second airport. This is policed under a contract between the Sussex force
and the airport authority. It supports a special police unit – heavily armed
and, nowadays, trained in counter-terrorism techniques. The internal
contrasts of Sussex were illustrated by several of our interviewees, who
mentioned that car speeding was a real problem in the sleepy rural lanes of
Sussex, but hardly in the traffic jams of Brighton.
1968
Sussex Police formed from a merger of five previous forces. In the 1960s
the force had been autocratic and hierarchical, ‘very military’ (Interview
35 – ex-senior police officer). ‘We policed the community the way we
thought best. We were doing it to them’. As far as planning was concerned,
‘We had no objectives’ (interview 35).
What happened locally? Police 103
Gatwick Airport
EAST SUSSEX
WEST SUSSEX
(Police H.Q.)
Lewes Hastings
Hove
Brighton
ENGLISH CHANNEL
1975
Sex Discrimination Act. ‘It was a truly unpleasant time’, said one senior
woman police officer: ‘It had taken a long time for the fuss that started in
1975 to die down.’ Previously woman police officers had only been allowed
to work in units commanded by other women, and they could only work bet-
ween 0900 and 2200. The changes triggered all sorts of cultural resistance,
including nasty letters to the local Brighton paper, which had been pinned
up in the (interviewee’s) police station. ‘We barely tolerated women’ (inter-
view 35 – ex-senior police officer; confirmed by several other interviewees).
1980
Chief Constable George Terry complained that: ‘In this day and age,
although it is understandable, almost everything is being measured in
money! I am, however, somewhat at a loss to understand how one meas-
ures peace of mind in terms of cash . . .’ (Sussex Police, 1980, p. i). He also
mentioned as worrying: (1) the increase in the number of burglaries; (2)
drink driving; and (3) the increase in the work of the Operations Room,
especially the rapid rise in emergency calls (p. ii). Force strength was 2823
on 31 December 1980.
1981
The Chief Constable in his annual report referred to his continuing
efforts to direct more of his officers to foot patrol: ‘The result was the
104 Continuity and change in public policy and management
1982
The Chief Constable (CC) pointed out that in 1982, 62 025 crimes were
reported compared with 29 124 in 1968 (a rise of 113 per cent). Over the
same period manpower had risen from 2385 to 2822 (20 per cent).
The CC remarked that: ‘The keynote for the future must indeed be the
prevention of crime and to this end the help of the public as a whole is abso-
lutely essential along with the inter-relationship of all law-enforcement
agencies.’ But he went on to say: ‘The relationship between the police and
the public should not be confused or directed solely towards the extreme
development of what are being called consultative committees, otherwise
our vociferous elements will wish to spend their time talking incessantly
and hindering the preventative role of the police’ (Sussex Police, 1982, p.
4). (As we saw in Chapter 2, this was the time when left-wing elements on
the Greater London Council were attempting to set up Council commit-
tees that would monitor the police. This was regarded with great suspicion
by many police, although by the end of our period it had become quite
routine.)
After three years of ‘nil’ growth all areas of the budget have been reviewed,
pared and constantly monitored until on reaching the present stage there is
no room left for manoeuvre. Since 77% of the total budget figure relates to
salaries, it is difficult to see how it will be possible in the immediate future for
the service to further develop with improved techniques, in order adequately
to support the most important part of the service, the uniformed patrol officer.
(p. 15).
1983
Roger Birch took over as Chief Constable, and opened his annual report
by stating that: ‘It is a well established fact that the style of policing unique
to this country depends for its efficacy almost entirely on its acceptance
and consent by the local community’ (Sussex Police, 1983, p. 3). He sug-
gested that changes recommended by the Scarman Report (see Chapter
4) were not necessary in Sussex, and gave his opinion that: ‘there is no
evidence to suggest a need for a major change in the area of consultation
and . . . to introduce change for change’s sake can be counter-productive’
(p. 4).
What happened locally? Police 105
1984
The Sussex Police Annual Report 1984 referred to both ‘the substantial
and protracted role played by the Sussex Police in providing their share
of aid to hard-pressed Chief Constables in the mining areas’ (p. 3) and
‘the trauma of the terrorist attack on the Grand Hotel, Brighton, which
became instant world news and which tested the professional skills and
morale of the Force as never before’.
The bomb attack (carried out by the IRA against government leaders
in Brighton for the Conservative Party Conference) took place at 0245 on
12 October. Senior politicians were killed and injured, and Mrs Thatcher
herself only just escaped. The Grand Hotel bombing became a leading
news item worldwide. In the search for evidence 3790 dustbins and 36
skips of debris were recovered and removed, in poor weather, and 40 000
vehicles daily were diverted while the site was secured and investigated.
The annual report also mentioned that the Police and Criminal Evidence
Act would require extensive training for every officer, and the extension
of the fixed penalty system would also require more training – as well as
adding an administrative burden.
The report refers to Home Office Circular 114/83, Manpower, effective-
ness and efficiency in the police service. The annual report says ‘Perhaps more
than any other of recent times, this circular has concentrated the minds of
senior and middle management levels in the service’ (p. 50). Emphasizes the
importance of timely and accurate management information.
Offences up 9.1 per cent to 63 971. Detection down 1.2 per cent to 38.3
per cent.
1985
The Chief Constable opened by mentioning the importance of com-
munity links, and singled out ‘the growing enthusiasm of the public for
Neighbourhood Watch’ (Sussex Police, 1985, p. 2)
He also mentioned that an investigation carried out into the adequacy of
the policing arrangements for the Conservative Party Conference exoner-
ated the Sussex force from all blame, adding that: ‘The implications for the
future in terms of manpower and equipment are enormous but there can be
no stepping back from the terrorist threat, which will not go away’ (p. 2). (It
is interesting to find statements like this a decade and a half before the 9/11
attacks.)
Later sections of the annual report show how other demands can under-
mine the intention to emphasize traditional, ‘beat’ policing:
and searches for a period of some three weeks. During that time there were 223
suspect packages and hoax bomb calls in the Division, all of which were inves-
tigated by members of that group. To help form such a team it was necessary
temporarily to withdraw most of the Resident Constables in Brighton. Such
was the measure of public feeling that many residents in the town wrote to the
Chief Constable expressing their concern over that move.
1986
The annual report mentioned that one aspect in which Sussex was ‘ahead
of the field’ was:
(This is, of course, exactly the kind of development which takes officers
away from traditional policing.)
Offences were up by 0.07 per cent and detections down by 15.8 per cent.
Government grant constituted 42.5 per cent of income, and constituent
authorities 41.5 per cent.
1987
The CC’s foreword to the 1987 annual report began:
In writing this foreword I have concentrated at the expense of all else on a single
subject which has far reaching consequences for the future well being of East
and West Sussex, namely police manpower.
It is a stark fact that since 1979 the authorized establishment of the Sussex
Police will have been increased by a mere 12 officers, if the special needs of
Gatwick Airport are set aside’ (Sussex Police, 1987, p. 2).
It went on to ask for ‘a totally new approach to the way in which police
establishments are determined and a willingness on the part of local and
central government to grasp the nettle of financial implications’ (p. 2).
For the first time this report contained short sections written by officers
What happened locally? Police 107
and civilian staff. The first, by Inspector George Divall, said: ‘On starting
at Shoreham I realized how uniform policing and policemen had changed.
The job itself had become violent.’
3700 Neighbourhood Watch (resident participation) schemes were
now operating in Sussex: ‘The continued growth of schemes has inevita-
bly created a problem in providing effective support from limited police
resources’ (p. 31). So Neighbourhood Watch assistants had been intro-
duced on an experimental basis.
1987 was also the year in which the local authority in Brighton created
a special committee for monitoring the police. A new advisor came from
London who had had experience with the Greater London Council (GLC)
system (see Chapter 4.2). Although this was politically a sensitive develop-
ment, relations between the Sussex Police and this new committee never
became as hostile as they had in the GLC (several interviews).
1988
The annual report opened with: ‘Chief Constables cannot hope for addi-
tional police manpower unless they are able to demonstrate that they have
gone as far down the road of civilianization as possible’ (Sussex Police,
1988, p. 2).
The 1988 report included a ‘Force Statement of Purpose’, which said
that:
The aim of the Sussex Police is to contribute, in co-operation with other agen-
cies in the community, to the development of the quality of life in Sussex by pre-
serving a peaceful society, assisting those in need, protecting life and property
and preventing and detecting offences. Sussex Police exists to provide a caring
service to the public.
The total establishment numbered 2923, and 56 per cent of the Force
budget went on police salaries, 12 per cent on pensions, 9 per cent on civil-
ian salaries, 5 per cent on National Insurance.
1989
Chief Constable began his report by noting that: ‘Sadly, many of the
headlines have reflected alleged police impropriety or apparent inefficiency
rather than success’ (Sussex Police 1989, p. 2). But he went on to say: ‘I
happen to believe that we are equal to, if not better than, a large part of the
private sector when it comes to the management of resources and that we
are streets ahead of the majority in the business of managing people’ (p. 2).
Commenting on the relations between the police, local authorities and
central government, the CC said:
108 Continuity and change in public policy and management
There are already signs that the balance of the tripartite arrangement is tilting
towards the centre, the latest illustration of this being a move to the central
control of capital expenditure which could well damage the initiatives of
forward thinking and progressive Police Authorities such as Sussex. (Sussex
Police, 1989, p. 3)
1990
In the annual report the CC described the research project Policing in
Sussex. A team from the University of Sheffield had designed a survey of
public opinion; 646 people were interviewed, 341 of whom were randomly
selected and the other 305 had been in contact with the police during the
previous year. When asked about problems in their areas 57 per cent cited
at least one, including traffic, noise, dogs, congregating youths and van-
dalism. Crime was rarely mentioned spontaneously, but burglary, vandal-
ism and theft were seen as worrying. Only 27 per cent admitted a strong
fear of crime, and this appeared to be affected by whether there was a com-
munity beat officer for the area and whether police were seen patrolling.
‘There was overwhelming support, more than 85% for the concept of local
officers, who were more clearly visible in the rural areas than in the towns’
(Sussex Police, 1990, p. 11).
On 30 July Ian Gow, Conservative MP for Eastbourne, was killed
by an IRA bomb which had been placed under his car at his home near
Eastbourne.
1992
The last annual report of CC Sir Roger Birch. He used it to reflect on the
changes after 39 years in the police:
The pace of life has quickened, moral standards have changed radically and the
family bond has loosened. Society is restless and changing, at an accelerating
pace but in an uncertain direction. Crime has escalated and knows no bounda-
ries, national or international. The cancer of drugs is spreading and publicity
What happened locally? Police 109
given to mindless violence induces fear in even the most peaceful community.
The phenomenal increase in traffic has added enormously to the police task.
(Sussex Police, 1992, p. 3)
1993
The first annual report with Paul Whitehouse as CC. He wrote that:
As the changes facing the police service nationally take effect Sussex Police,
like every other Force, will increasingly be judged on its performance. There is
thus a danger that our activities will concentrate too much on achievement in
the short term. We need to devise long term plans to balance this effect. (Sussex
Police, 1993, p. 2)
The report recorded that, after a major consultative exercise with the
Police Authority and community groups, the Sussex Police Standard had
been published. This ‘landmark document’ ‘spelt out to the public the
standard of service they properly expect from Sussex Police in each of
those areas’ (the areas were: response to calls, crime, public order and reas-
surance, traffic and community assistance).
1994/95
(From now on the annual reports followed the financial year rather than
the calendar year.)
The CC reported that: ‘The adoption during 1994 of our Force Crime
Strategy enabled us to capitalize on the lessons learned during our anti-
burglary initiative, Operation Bumblebee, launched the previous October.
This was based on intelligence and actively targeting active criminals’
(Sussex Police, 1995, p. 2). He also mentioned animal rights demonstra-
tions at the port of Shoreham, where livestock exporters were operating.
A major police operation had been required. The first Local Policing Plan
for Sussex was now ready.
The CC also remarked that: ‘I have made it clear that we will not allow
our integrity to be compromised by the pressure to improve our position
in national league tables’ (p. 3).
The main body of the report gave details of the Force restructuring:
The aim was to identify police resources more clearly with the community so
that policing can become more truly accountable. For this reason parishes were
110 Continuity and change in public policy and management
grouped into natural policing sectors. These sectors are to be the key units for
the delivery of police services. (p. 6)
The Sussex Police Standard, published in 1993, set out the minimum levels of
service the Sussex public is entitled to expect in five core areas of policing –
crime, response to calls, public order and reassurance, traffic and community
assistance. The Force Service Plan, and subsequently the Policing Plan for
Sussex, set out targets against which to measure our performance. These targets
have been set by the Chief Constable in consultation with the Sussex Police
Authority and in response to the Key Policing Objectives set nationally by the
Home Secretary. (p. 7)
13 August 1994: An IRA bomb exploded in Bognor Regis. Later the same
day a larger bomb was found, unexploded, on the seafront in Brighton.
1995 saw a huge police operation at the port of Shoreham, just west of
Brighton and Hove. Animal rights protesters attempted to prevent trucks
carrying live animals for export from reaching the waiting ship. At first
the police were surprised – 100 officers failed to control the situation.
Subsequently up to 1300 officers were involved, with reinforcement being
drawn from neighbouring police forces. The bill for police action soon
exceeded €4 million. Eventually the police decided to limit their policing
effort, which led to a partial ban on the controversial exports, and to court
actions against the police.
At the same time (April) the CC found himself defending the low place
Sussex had achieved in a national league table on crime detection. Paul
Whitehouse argued that the league tables were calculated in a misleading
way (The Argus, 1995, p. 1).
Also in 1995, the method of fixing the central government grant to police
authorities changed. ‘In practice the new system gave the Department of
the Environment [local government ministry] a more influential role over
police finances’ (interview 4 – ex-senior police officer). In the same year
police authorities were restructured (see Chapter 4.2 – 1994 Police and
Magistrates’ courts Act).
1997
The government agreed to pay €750 000 towards the cost of policing the
Labour Party Conference in Brighton. Both major parties frequently held
their annual conferences in Brighton and the cost of the extra security had
hitherto fallen on Sussex taxpayers.
1998
In January armed police in Hastings shot dead James Ashley, while car-
rying out a drugs raid. It subsequently transpired that Ashley had been
What happened locally? Police 111
unarmed, and had not been one of the persons the police had been looking
for. The incident, which became known as the ‘Hastings Shooting’ was
immediately referred to the Police Complaints Authority and also became
the subject of an independent investigation by senior police officers from
outside the Sussex force. The Hastings Shooting became a major and
ongoing media story. As it unravelled over the next three years six officers
were suspended (including a Deputy Chief Constable) and eventually, in
2001, Chief Constable Paul Whitehouse took retirement. The aftermath
of the shooting ‘tore the superintending ranks apart. Everyone took sides
. . . At that point we began to be a failing organization’ (interview 13 –
senior police officer). ‘Massive ramifications’ (interview 29 – senior police
officer). It was ‘a real watershed – we learned that if it wasn’t written down
it hadn’t happened’ (interview 14 – senior police officer)
2001
In July the merger of the Brighton and Hove police districts into a single
division was approved. This followed the award of city status to Brighton
and Hove combined.
After unprecedented pressure on the Sussex Police Authority from
the Home Secretary (David Blunkett) over the Hastings Shooting, Chief
Constable Paul Whitehouse gave notice ‘with some sadness’ of his inten-
tion to retire: ‘my priority now is to protect the good name of the Sussex
Police’ (Patrol, August 2001, p. 3). ‘We lost a lot when he left – he had
encouraged a more open environment, more consultation, a more ana-
lytic, questioning approach’ (interview 16 – ex-senior police officer, and
other interviews). Deputy Chief Constable Maria Wallis took over tem-
porary operational command of the Sussex force, until Ken Jones was
appointed Chief Constable in 2002.
2002
Ken Jones – the new CC – pushed for visible up-front accountability and
immediately launched a full-scale operational review. ‘League tables have
become much more important’, performance indicators were vigorously
applied to Basic Command Units (BCUs) and the Chief Constable would
hold regular meetings interrogating his officers about their results (inter-
views 14 and 11). ‘We’ve got to be a 24 hour service and achieve results.
Performance has become almost your raison d’etre’ (interview 11).
2002–03
The Joint Annual Report of the Sussex Police Authority and the Chief
Constable of Sussex Police explained that, following a 15-month opera-
tional review, 42 Inspector-led neighbourhood police teams had now been
112 Continuity and change in public policy and management
set up covering every part of Sussex. (Above that there were 12 Districts,
each matching local authority areas and commanded by a Chief Inspector.
Brighton and Hove consisted of three Districts.) It also mentioned 62 new
Police Community Support Officers (PCSOs) who would be patrolling the
streets. Crime fell again – 600 fewer recorded crimes than two years pre-
viously (3.1 per cent). The number of crimes detected increased by 5 per
cent. The report showed detailed achievements in different aspects of crime
against targets (for example, ‘We aimed to reduce the number of domestic
burglaries by 2% and increase the burglary detection rate to 18% . . . There
was a 7.8% increase in burglaries and a detection rate of 12%’ (p. 6).
48 per cent of funding came from Home Office grants, 10 per cent from
Revenue Support Grants, 28 per cent from local Council Tax and 17 per
cent from non-domestic rates.
2003–04
That there were two public consultation exercises each year, and the two
top priorities for the public were: (1) more policing on the streets; and (2)
better call-handling arrangements (Sussex Police Authority, 2004a, p. 2).
The report said that key crime indicators, including vehicle crime, were
moving downwards, but anti-social behaviour had increased and was ‘a
key issue for policing in Sussex’ (p. 4). More than 200 Police Community
Support Officers were now deployed to help tackle this, ‘helping to keep
a lid on low level incidents – loutish behaviour, vandalism and the like’
(p. 4).
The Annual Report included a section on an independent assessment of
Sussex Police by Her Majesty’s Inspectorate of Constabulary (HMIC). It
said Sussex was rated as excellent in 3 areas, good in 7, fair in 5 and poor
in only one – call handling.
2004–05
The annual report congratulated itself on having had Neighbourhood
Policing Teams in place since 2003 (the government had just announced
that it wanted them throughout the country by 2008) (Sussex Police
Authority, 2005, p. 1).
The detection rate was up from 24.2 per cent to 25.4 per cent. Burglary
(–13.5 per cent), robbery (–13.5 per cent), car crime (–19.1 per cent) and
What happened locally? Police 113
The above timeline should have given a flavour of the issues confronting
the Sussex police over the four decades from 1965 to 2005. In this analyti-
cal section we will draw out some of the more prominent themes.
First, the impacts of central government initiatives were considerable
throughout the period, but the frequency and detail of these initiatives
seems to have grown since the late 1980s or early 1990s. ‘What is happen-
ing now is that central government is coming up with something new every
week or so, and we are going round the change cycle much more quickly’
(interview 13 – senior police officer). PACE was a watershed in 1984, but
the process accelerated further from the early 1990s. Interviewee 29 saw
the 1993 Sheehy Report (Inquiry into Police Responsibilities and Rewards)
as a milestone (Sheehy, 1993).
Interviewee 28 argued that the balance of power between the police, the
Police Authority and the Home Office ‘has shifted hugely’ – in favour of
the Home Office. ‘Now they have a huge hand on local policy – it is almost
delivered to a formula’ (interview 16 – ex-senior police officer). ‘The inter-
ference is more. The ability of Chief Constables and the BCU commanders
to deliver a locally-tuned service is better than before [but] you’ll never get
away from a targeted, performance-driven culture’ (interview 14 – senior
police officer). ‘There has been an increasing emphasis on central targets’
(interview 2). Some, while acknowledging the growth of central controls,
argued for a more nuanced picture: overall ‘the Chief Constable still has
a lot of room for maneouvre’ (interview 4 – ex-senior police officer). In
Sussex, for example, they resisted the national trend to use CS gas (prefer-
ring pepper sprays) and they went against the national norm by employing
114 Continuity and change in public policy and management
cannot do all that people would like us to do, all the time. When I, or any
of my colleagues, have to choose between competing demands for our
service, the priorities contained within the Performance Plan are our guide’
(CC Paul Whitehouse, 2001, p. 5). This is an interesting issue in terms of
change because it leads to many small, short-term changes (Tortoise-like)
but in the longer term looks more like an alternation between two desir-
able but mutually exclusive poles.
Third, it is absolutely clear that, as was the case with hospitals (Chapter
3) the salience of management, of a managerial way of thinking and
organizing, has grown enormously over the period since 1975. Most of
our police interviewees picked this out in one way or another. Interviewee
29 said that notions of good management had become very prominent in
police training from the late 1980s. Interviewee 35 referred to the Home
Office Circular 114 of 1983, which focused on efficiency and effectiveness:
‘previously we hadn’t managed resources at all’. Many interviewees com-
mented on the growing impact of performance indicators and targets,
beginning in the mid-1990s but reaching a new peak in Sussex after Ken
Jones became Chief Constable in 2002. One of the interesting differences
between the police story and the hospital story, however, is that, by and
large, the police have remained as their own managers. In the National
Health Service (NHS), as we saw in Chapter 3, there has been a large
growth in professional managers who have no clinical background, but
nevertheless take responsibility for running hospitals (including, since
1997, responsibility for the clinical performance of hospitals – that is, what
doctors and nurses do). The police, however, have not placed themselves
under a new cadre of secular, non-police managers. They have trained
themselves to be modern-style, performance-oriented professionals and
leaders. Of course they always did have a clear management style of their
own – highly disciplined and hierarchical – in a way that the clinical pro-
fessions arguably did not, and it seems that they have been able to build
upon and adapt this. This assumption of a far more managerial approach
has been felt throughout, but has impacted differently on different ranks.
Several of our interviewees, when asked what they thought had not
changed so much, mentioned that the work of a constable on the beat
had considerable elements of continuity to it – the situations they were
confronted with, the values, the camaraderie, and so on. Even more fun-
damentally, ‘they are still a disciplined, hierarchical force’ (interview 2).
However, our interviewees immediately went on to say that for the super-
intending ranks the job had changed enormously, principally because of
the range of management responsibilities and considerations which now
defined the lives of those more senior officers. This seems a reasonably
clear case of a Stalactite.
116 Continuity and change in public policy and management
WILSELE/WIJGMAAL
13.84
13.69
5.99 KESSEL-LO
LEUVEN
23.98
HEVERLEE
Figure 7.2 Map of the City of Leuven (after 1977, with the km2 of the
districts)
Gendarmerie
2359
Districts Municipal
and related
local police
Note: The figure should not be read as implying that the municipal police were
hierarchically subordinate to the Gendarmerie. Also, this figure does not show either the
judicial police or the specialized transport forces.
Gendarmerie
589
Districts Municipal
Police
Federal Police
196 Local
Police
police zones. This was the local outcome of the fundamental national
reorganization described in Chapter 4. The Leuven Rijkswacht District
covered a considerably larger area – about 500 000 people, with 13 bri-
gades and 50 BOB agents (total about 400 people – BOB (Bijzondere
Opsporings Brigade) is the Rijkswacht’s guard and investigation brigade
– a kind of Criminal Investigation Department, CID in English terms).
However, within Leuven city, the Rijkswacht was complementary to
and non-competitive with the municipal force. Most of the Rijkswacht
support was for smaller villages to guarantee a 24/7 service. Community
policing became one of the guiding principles for local police. In order
to improve the collaboration between Rijkswacht and Municipal Police
services, Interpolice Zones had been established from 1995 onwards (see
PZ Leuven, 2006). This was beneficial for Municipal Police stations since
they could transfer night shifts to the local Rijkswacht. The 1997 Act
allowed mayors to have a say in the local districts of the Rijkswacht. It
was hoped that police activities would therefore become better matched
to local needs.
The Royal Decree of 7 September 2001 established the local police basic
functions within the spirit of community-oriented policing. However, the
structure of a local police organization is not fixed by statute. The six
basic functions were district operations, reception, intervention, caring
for victims, local investigations, and maintaining law and order. In addi-
tion to these six basic functions, local police could also be assigned tasks
from the federal level. District operations imply that police are visible,
approachable and responsive, given local circumstances and density.
Reception implies that citizens always should have the opportunity to
contact police services, intervention implies a proportional response to a
call. Caring for victims should result in adequate reception, information
and support provided to victims (although, in general Belgium has a lower
level of such support than does the UK – see Chapter 5, Table 5.4). For
120 Continuity and change in public policy and management
The origin of the Leuven police was related to the fire service, which had
first been set up in 1807. With the foundation of Belgium in 1830 the
municipal government confirmed the complete integration of the munici-
pal police and the fire service. In 1841 the city established a municipal
neighborhood watch, initially volunteer-based but very soon professional.
The city consisted of 60 geographical sections. Public criticism of the
organization of the police function and its equipment continued. In 1879
the fire service was replaced by a new police organization consisting of a
Commissaire, eight Adjoints, and 30 policemen for six districts of the city.
One of their responsibilities continued to be the fire service. Social and
student riots in 1901 and 1902 resulted in further criticism of the police.
During the First World War, initially the Germans abolished the police,
but then re-established it. In 1924 white helmets were made mandatory in
order to make policemen more visible for cars. In 1935 a preventive air
attack watch service was established. This eventually led to the separation
of the police and fire services. 1942 marked a key turning point for the
municipal police since recruitment and structured training was organized
for the whole country. Also the strength of the force was determined per
municipality. For Leuven this was 87 officers, for Kessel-Lo the number
was 17. The German occupiers were very critical of the police and accused
them of political crimes and acts against the occupying authorities.
Leuven was liberated on 1 September 1944. Police Commissaire François
Chevalier, who was in office from 1935 till 1969, was accused of collabora-
tion with the Germans, but he was subsequently recognized as being part
of the armed resistance.
The schools for officers (Tervuren) and other ranks (Bokrijk) which had
What happened locally? Police 121
been established during the war were abolished. They were re-established
by the provincial council of Braibant in 1948. This bilingual school for
officers was split in 1966. For ordinary police men the re-establishment
of the police school with compulsory training did not take place until
1986. However, the Leuven Police were obliged to take classes and train-
ing from 1971, in the Brussels police school, on the initiative of the new
Commissaire who was originally from the Brussels Police. Until 1975,
Leuven Police, when hired, had to prove that they were able to provide
information in French (even though this was now an illegal requirement
according to the language laws).
1968
The basis of an entente between Rijkswacht and the Leuven Police grew
from the ‘Leuven Vlaams’ crisis (de splitsing – discussed in the previous
chapter). The Rijkswacht guaranteed good public order. In fact, from
1968 until 1975 the Rijkswacht rather than the municipal force dominated
the city.
1971
Mayor Smets (Christian Democrat) was often absent and was replaced
by his First Alderman, Louis Tobback (Socialist). Ray Ieven became the
new head of the Leuven Police and succeeded François Chevalier who had
been chief of the Leuven Police from 1935 till 1969. Ieven remained head
of police until his retirement in 1995.
There was a shooting incident in which a student was killed. Ieven was
pressured to resign in the press and by students. He organized shooting
training, which had not previously existed.
1975
At this time there was a tradition, mainly among Francophone students,
to have regular Thursday riots in the centre of the city:
It became part of the folklore. The Gendarmerie had an agreement with the
students on this issue. Slowly this folklore disappeared: the culture of students
changed, the political polarisation disappeared (between left and right), and ‘de
splitsing’ pushed out the francophones. (interview 31)
1972
Decision to modernize the car fleet and to create civilian assistants for the
uniformed police officers.
1976–77
Merger of the Municipal Police forces of Leuven, Heverlee, Kessel-Lo and
Wilsele. Ievens developed a comprehensive and technical document for
the reorganization of these four entities (Ieven, 1976). Within the merged
organization new specializations were established. Alongside the regular
administrative services and surveillance with district services, traffic, judi-
cial and social services were established. Later on a dog section and motor
cycle section were established.
Mayor Smets reached the end of his term of office. He had been Mayor
for the period 1947–52, and then for another three terms from 1958 until
1976. Alfred Van Sina, up to 1976 the Mayor of the previously separate
village of Kessel-Lo, became Mayor of Leuven, where he remained for 18
years, until the elections in 1994.
The mergers of the municipal administrations in general, and of the four
police organizations in particular, were not easy. Four to five years later ‘the
1976 merger was still not processed by the organisation’ (interview 26).
After the merger, the situation remained political and parochial:
If you have a party membership and a uniform you are a police man. All con-
stituent parts of the police (previous municipalities) were autonomous: Van
Sina (mayor of Kessel-Lo) protected ‘his’ previous police men. There was also
serious understaffing. About 185 people were available but 220 were needed.
(interview 27)
1980
The first female police officer was appointed.
1982
Mayor Van Sina was re-elected and started his second term of six years.
1988
Mayor Van Sina was re-elected and started his third term of six years.
1989
Louis Tobback became Minister of the Interior (until 1994). One of the
members of his cabinet was Lode De Witte, who later became Tobback’s
Chef de Cabinet (Chief of Staff). Later, in 1995, De Witte became
Governor of the Province of Flemish-Braibant, with extensive police
responsibilities.
What happened locally? Police 123
1992
Implementation of the 1992 Police Function Act describing the respec-
tive competencies of the Municipal Police, the coordination of tasks and
assignments of the three existing police forces (Rijkswacht, Judicial and
Municipal) through the ‘Pentagon’ platform.
Several initiatives were taken to encourage integration. This integration
policy was a combination of top-down policy and bottom-up voluntarism.
In the Leuven case there were special circumstances. Leuven was one of the
first pilots for all these initiatives. The Leuven IPZ (Inter Police Zone) was
guided by the Ministry of the Interior. Municipalities got financial incen-
tives to collaborate on a voluntary basis with other municipalities (in these
zones). For the Rijkswacht, also guided by the Ministry, it was compulsory
to collaborate and support these voluntary initiatives. The major objective
was to move from an IPZ status to become an integrated Police Zone (PZ).
1994
Louis Tobback was elected Mayor of Leuven. He took over after Van
Sina’s 18 years in office.
The management consultancy, Team Consult, was contracted to under-
take a local evaluation.
Louis Tobback became President of the Flemish Socialists (until 1998).
1995
Team Consult reported. They recommended better development of decen-
tralized district teams and a ‘holistic approach’; 76 of the 138 officers were
to be assigned to districts (wijkteams) and 62 to more centralized interven-
tion functions. The district teams were to be the principal ‘homes’ of offic-
ers, with periodic assignments to the intervention units.
Ray Ieven: retired after having headed the Leuven police for 24
years.
Hugo Michiels was Ieven’s successor and he started a reorganization
(‘fundamental contribution to the general security of the city of Leuven’).
He remained in office for nine years.
Lode De Witte, the former Chef de Cabinet of Tobback at the Ministry
of the Interior, became Governor of Flemish-Braibant with a special
emphasis on police and security policy.
As of September 1996 there were 215 police plus 25 civilians.
1996
Four district bureaus were established: Leuven Centre, Heverlee,
Kessel-Lo, Wilsele-Wijgmaal. An agreement on cooperation was reached
with the local district of the Rijkswacht (PZ Leuven, 2006).
124 Continuity and change in public policy and management
1997
Leuven City Security Charter (Veiligheidscharter). This was the opera-
tional consequence of the 1992 law. It identified certain priorites (kracht-
lijnen) for Leuven: burglaries and robberies, transport safety, car crime
and drugs (IPZ Leuven, 1997).
1998
Louis Tobback became Vice-Prime Minister and Minister of the Interior.
2000
Leuven became a Pilot Police Zone PPZ.
Mayor Tobback was re-elected for a second term of six years.
2001
Leuven Police Zone was created. The merger of the local branch of the
Rijkswacht with the Municipal Police meant that two-thirds of the Leuven
Rijkswacht staff went to Brussels and the others (about 55 officers) joined
the Leuven Police (which by then consisted of about 220 officers, and
about 280 total staff, including civilians).
‘There was a perception that the Gendarmerie was absorbed by the
police’ (interview 27). In practice the new organization was designed and
personnel needs were determined. All staff had to apply for the new posi-
tions internally. About 90 per cent got their desired positions.
2003
Physical co-location of Leuven Police HQ and the local branch of the
Rijkswacht, and relocation of the Leuven Police Zone in new buildings on
the Phillips Site.
2005
The Zonaal Veiligheidsplan 2005–2008 (local security plan) set out the
general context, the contributions from the local police, the resources,
and the special projects (PZ Leuven, 2005). This presented quite a detailed
statement of planned inputs and standards (for example shift patterns for
the police on intervention duties; numbers of district officers in relation to
resident populations).
2006
Louis Tobback was re-elected as mayor for a third term of six years (when
his term expires in 2012 he will have served for 18 years as Mayor).
Michiels resigned as Chief of the Leuven Police. He had served for nine
years.
What happened locally? Police 125
2007
Jean-Paul Mouchaers was appointed as new Korpschef. He quickly devel-
oped a diagnosis of the problems facing his organization. He wanted to
use processes as the main basis for the organization rather than vertical
structures. The presence of former Rijkswacht staff was not considered to
be a significant influence on the culture. The old cultures were said to be
disappearing, partly because there was a high turnover of personnel. There
were plans to consult the population (target groups, stakeholders) in
forming the elements for a new local security plan. There was a refining of
the existing security monitoring process by adding new measures of public
perceptions of insecurity. New tasks made their call on police resources
and skills, for example the Prosecutor-Generals asked the police to focus
on intra-family violence (two directives from the College of Prosecutors
General). And there were plans to organize a satisfaction survey among
the personnel.
The published monitoring report was still mainly limited to budgetary
data. It did not yet include much non-financial management information,
particularly with respect to performance outputs or outcomes. The per-
ception was that external communication and information and commu-
nication technology (ICT) needed to improve (including the website and
the yearly report). There was a need to think in the long run by combining
systems of calls, with despatching, geographic information system (GIS)
and closed-circuit television (CCTV). This would allow faster and better
reactions.
Many models are put forward to make things happen: Initially there was BPZ
[Basis Politie Zorg: Basic Police Care] which was translated into six basic
functions. Then there is the model of ‘Excellent Organisation’ with EFQM.
Now there is IGP [Informatie Gestuurde Politie: Information Guided Police].
(interview 26)
that new models arrive before the previous ones have been implemented, that
all these models are top-down (from Brussels), that the consistency of all initia-
tives is lost, that there is insufficient training. Is there an impact on change, and
is this improvement? There is a loss of expertise: nobody knows what a district
cob is [wijkagent, or local constable]. Informing police actions is not just about
numbers, it is about district police transferring their intelligence to other parts
in the organization. (interview 26)
Doubts of this kind are certainly not confined to the Leuven Police. Here
the Chief Commissioner of another Belgian force:
Despite these anxieties, however, there was a clear agenda that the police
needed to be upgraded to professional and modern standards. This
included training and quality.
At one time competition between the different police forces was endemic.
For example, the Rijkswacht had developed a quality policy around
BPZK (Basic Police Care with Quality/BasisPolitieZorgKualiteit) This
was a bottom-up type of initiative which included elements such as :
● reception desks;
● involving policemen in the process of improvement;
● abandoning the traditional model that ‘officers think, police consta-
bles execute’;
● recontacting victims;
● shopping days within the police: a ‘market’ for police projects;
● defining basic police care.
In both cases the Leuven Police claimed the authority to perform these
functions; Commissioner Michiels had said that there should be no
campus bureau for the Rijkswacht, the students were ‘ours’. This was a
competitive claim (interview 30).
In the earlier part of our period the position of the Rijkswacht was:
● the smaller the local police zones the better (this would strengthen
the federal police);
● divide and rule;
● central police has more power when local police are dispersed.
128 Continuity and change in public policy and management
The pressure to move to bigger local units came from the Municipal Police
themselves. The mayors were usually united in favour of smaller units.
Eventually, however, a measure of cooperation was achieved. A
good Pentagon platform was established, consisting of the Mayor, the
Commander of the Rijkswacht, the Municipal Police, the Judicial Police,
and the Royal Prosecutor. Their choice was for a one-city police zone in
Leuven. There were several reasons for this decision:
● some mayors of the smaller municipalities did not want to join the
bigger unit of Leuven (for example Herent);
● there was a tendency within the province to go for smaller rather
than larger – there was some disagreement in several smaller munici-
palities between the Chef de Corps of police and the Mayor.
Tobback (SP) Vande Lanotte (SP) Tobback (SP) Van den Bossche (SP)
1988–94 1994–98 1998 1998–99
Mayors of Leuven
This was therefore a story of strong local mayors and police chiefs,
embedded in strong elite networks. They had strong connections and
ensured participation of local elites in national decision-making. They
managed to secure funds, influence mergers (and so on), to the benefit
of their communities and organizations. This pattern therefore contrasts
with the experience in Sussex. Whilst it is true that most of the Sussex
chief constables remained in office for more than five years, their average
length of service was considerably less than that for Leuven police chiefs.
More significantly, perhaps, for them and for the local politicians, the
national government level was more ‘distant’ (‘them and us’) in the Sussex
case. Furthermore the demands and interventions from central govern-
ment were more volatile, as different governments and different Home
Secretaries came and went.
There were important developments in the Leuven story, but also consider-
able continuities. First, the changes. Obviously, the relationship with the
national force changed considerably, largely in the direction of somewhat
greater equality and more cooperation. Connected to that, the Pentagon
(five-cornered) system of coordination between the Municipal Police, the
Judicial Police, the Rijkswacht, the Mayor and the Prosecutor became
an important new way of working. And, as in Sussex, each Police Chief
(Korpschef ) brought with him a new style and mixture of skills. There were
also bottom-up changes such as the dissatisfaction by local politicians which
resulted in a local evaluation in 1994–5, and the fact that in 1999 Leuven
volunteered as a pilot project for the merger of Rijkswacht and Municipal
Police (there were only five pilots in the whole country). Obviously the fact
that the responsible Minister was Tobback – also the Mayor – is important.
Nor should we underestimate the change-promoting effects of external
events. As one of our most experienced interviewees said: ‘Catastrophies
made a convergence of concepts. It is important to know where to go. One
needs to be ready when needed’ (interview 34).
Then we come to the continuities:
(Although one might question the first part of this claim, given that
student riots were at one time commonplace and regular.)
The two local police stories can be told in different ways. The contrasts
are perhaps more ‘front stage’, but just behind them lie some important
similarities.
The three most obvious contrasts are, first, the different structures and
scales of police forces, second, the distinctive role of local politicians in
the Belgian case and, third, the far greater availability of performance
information in the Sussex force. The first contrast can be stated bluntly
by saying that, for the people of Brighton, there is and has been only one
police force, whereas for the people of the Leuven area there have been
many (see Figure 7.2) and even now, after all the structural reforms and
mergers of the past three decades, there are still two quite distinct bodies
– the Leuven Police and the Federal Police. To take the point further, the
Leuven Police as a whole are considerably smaller than just the Brighton
and Hove division of the Sussex force. These differences in size have sig-
nificant implications for the range of specialist units and equipment which
a force can support, and for career patterns. In Belgium the specialist units
tended to be at national level and it was only after the 2001 merger of
municipal forces with the local Rijkswacht that some local forces experi-
enced an injection of highly trained professionalism.
As for the role of local politicians, the English police authorities do
contain (seconded) local politicians, but these bodies are seldom con-
sidered to be particularly powerful influences on chief constables. An
active and prominent member described them in the period before the
1994 Police and Magistrates Courts Act as ‘a backwater you put the old
stagers on’. The arrival of independent members after the Act produced
somewhat more activism (interviews 4, 8) but did not transform the situa-
tion. In the one case where the Sussex Police Authority appeared to take a
decisive executive action – securing the early retirement of Chief Constable
Whitehouse in 2001 – the Authority was evidently acting under consider-
able pressure from the Minister in Whitehall (interviews 8, 28). Usually
the Chief Constable ‘still had a lot of discretion and room for manoeuvre’
What happened locally? Police 131
(interview 4). The tightening-up of political control over the police has
(1990–2005) come almost entirely from the centre. Whitehall has laid
down an ever-tighter net of performance indicators, standards, plans
and ‘initiatives’. In Belgium, by contrast, local mayors were very much in
charge of the local police – not only in setting operational priorities but
even sometimes to the point of determining quite junior appointments
(interviews 26, 27, 30, 32, 34). This was doubly the case in Leuven, where
Mayor Tobback was (is): (1) long-serving; (2) also a national political
figure of considerable influence; and (3) particularly interested in police
reform, having developed a vision of his own as to the direction in which
reform should be heading. (We do not mean to imply that he intervened
in junior appointments – although some other mayors certainly did – but
simply that Tobback’s influence and experience were greater than his
counterparts in some other Belgian police zones.) Paradoxically, if the
Municipal Police were to gain greater professional autonomy, it had first
to be awarded to them by local politicians.
The third contrast – the much greater use of performance information
in the English force – shone out from almost every document we encoun-
tered. By the end of our period the Belgian documentation was certainly
reflecting managerial language, models and concepts, but the amount
of information about actual outputs and impacts remained very limited
by English standards. Most of the figures in the Belgian documentation
concerned more traditional inputs – how many officers were assigned here
or there, not what they had achieved. There were some figures indicat-
ing local crime trends, but on the whole these were not tied to specific
police activities, and certainly not to quantified targets. There was very
little comparative data. If one compares, for example, the Leuven local
security plan for 2005–08 (PZ Leuven, 2005) with the Local Policing Plan
for Sussex 2004–2005 (Sussex Police Authority, 2004b) the latter contains
quantified targets in every section and much more precise information
on where resources, both financial and personnel, are being allocated.
Nor, during our research, did we ever see in Belgium anything remotely
approaching the wealth of performance information contained in the
regular monthly Force Strategic Performance Information booklet which
is distributed to senior police officers and the officers and members of the
Sussex Police Authority.
We now turn to some of the similarities. We can mention at least three.
First, there is the move towards a more professional service, with more
explicit standards and better training (for example Bergmans, 2005). That
is evident in both localities, and emerges from most of our interviews with
senior police officers themselves. Second, there have been a series of tech-
nological changes which have various important effects. For example, the
132 Continuity and change in public policy and management
1. The 1984 Police and Criminal Evidence Act. This had obliged the
police to be far more systematic in handling suspects (interviews 1, 11,
16, 29).
2. The forced early retirement of Chief Constable Whitehouse in 2001.
This had proved traumatic for the whole Sussex force and had led to
What happened locally? Police 133
It can be seen that only the second of these – the Hastings Shooting and
the later resignation of the Chief Constable – can be construed as primarily
‘local’. All the others came principally from central government. And even
the Whitehouse resignation eventually involved strong pressures behind
the scenes from the Home Secretary. Issues 1, 3 and 4 are all genuine
policy changes – intendedly basic shifts with lasting effects.
For the Leuven force there was less of a consensus. On the one hand
almost everyone agreed that the 2001 restructuring was a watershed
(interviews 26, 27, 30, 31, 32). ‘Het werd een fundamentale ingreep in ons
politiebestel’ (It was a fundamental intervention in our police system) (PZ
Leuven, 2005, p. 1). But on the other hand, opinions differed about the
significance of some other issues. For example, the earlier (1977) merger
was seen as very important by some, but was not mentioned by others.
The Dutroux case and escape – a national event but one which affected
the reputation of the entire criminal justice system – was seen as impor-
tant by several respondants, but mainly because it gave the final impetus
for a further, deeper restructuring. The 1990 Pinksterenplan was clearly
acknowledged as a major step, but again it was a national-level initiative,
which provided an opportunity for movement locally. The pattern that did
emerge from the Leuven interviews was one of a relatively small group of
politicians who occupied a variety of roles at national, provincial and local
levels, and who pursued a particular direction of structural reform over
almost 20 years. As one insider put it: ‘it was a combination of personali-
ties, a long term vision which was not changing, and political opportuni-
ties’ (interview 32). Thus the pattern we see in the Leuven case is perhaps
one of change that takes longer but is driven by a more consistent political
vision, and one in which some of the key figures were based in or knowl-
edgeable about the particular circumstances of the locality. In the Sussex
case, however, we find a police force which is obliged to react to a series of
policies not at all of its own making. Things come down from the centre
and must be attended to. Of course, some of these things may be attractive
134 Continuity and change in public policy and management
to some senior officers in the local force, and even those which are not
popular will nevertheless receive a considered and disciplined response.
But there is more ‘distance’, more of a sense of ‘them and us’ than we find
in Leuven.
8. Reflections on theories of change
8.1 INTRODUCTION
The previous five chapters have revealed both similarities and differences in
policy and management in the two chosen services and countries. The differ-
ences are great – large enough for us to ditch any naive or mechanistic theo-
ries that ‘globalization’ or ‘reinvention’ or ‘the New Public Management’
are sweeping through the world, obliging every government urgently to
move in the same direction. Evidently even governments that are geo-
graphically adjacent, multiply interconnected and similarly subject to the
allegedly homogenizing effects of European Union (EU) membership can
behave very differently when it comes to policymaking and management.
The similarities between the countries and services are perhaps more
subtle, but no less profound. In rather brutal, bullet-point style we can
summarize the details of Chapters 3 to 7 as follows.
Main differences:
Main similarities:
135
136 Continuity and change in public policy and management
rhetoric and display and media attention around new policies in order to see
the big picture. That big picture usually includes other influences – cultural,
technological, even demographic – that may be simultaneously at work.
The above analysis is not meant to suggest that long-term organiza-
tional, technological and cultural trends are completely unconnected with
policies; that (for example) the former are never affected by policies, or
that policies are hatched in complete ignorance of or indifference towards
external trends. On the contrary, policies quite frequently engage directly
with these trends, seeking either to speed them up or slow them down (or in
some cases prevent them from proceeding altogether). Thus, for example,
there are policies to try to guide the types of medical equipment and speeds
at which new medical technologies are acquired by hospitals. These may
not be the most publicly and politically salient aspects of hospital policies,
but they are important nonetheless. Our point, therefore, is not that the
two types of change (policy or long-term external trends) are totally dis-
connected, but rather that, from a 40-year perspective, many policies come
and go, but the trends in organizational specialization, managerialization
and technological change look more consequential and less strategically or
tactically ‘chosen’ than are most policy shifts. In short, there is much more
than just policy to the stories of change in the hospital and police services.
A second initial observation would be that there does not seem to be
any obvious, regular, repeating or cumulative pattern to change, in either
country or either service. None of these stories are of a neat cycle of stability
followed by punctuation, followed by stability, followed by punctuation –
at least not with any regular periodicity. But neither are they tales of steady
incremental change, with each service moving step by step along a particu-
lar path. Patterns there may be, but of a messier, less predictable kind – we
will come back to them in a short while.
A third observation would be that the similarities and differences indi-
cated above seem to derive from different levels or locations in the overall
governance and social systems. (The following is not a perfect match, but a
rough generalization.) The differences appear to derive mainly from distinc-
tive, high-level, long-lasting features of the national political systems of the
two countries. Thus the faster pace of change in England and the greater
emphasis on quite ‘hard’ performance management were both outcomes
of the strong, majoritarian, ‘law-lite’, more centralized, political system in
that country, contrasting strongly with the weaker, federal, law-heavy, frag-
mented nature of government in Belgium. However, the similarities are gen-
erated by deeper, perhaps less obvious but extensively international trends in
society and technology. Thus the technological changes influencing hospital
design, clinical practice and, indeed, the cost of acute healthcare are interna-
tional and common to at least all advanced countries. They do not force a
138 Continuity and change in public policy and management
single solution in every case but they do seem to push hospitals in common
general directions. Something very similar can be said of police methods. Or
again, the increased emphasis on the need to consider public sector organiza-
tions of all kinds as managed entities is part of the huge growth of managerial
ideology and of management knowledge throughout the Western world and
beyond (Parker, 2002; Pollitt, 1993; Sahlin-Andersson and Engwall, 2002).
Fourth, the specific influences of EU membership in both these sectors
seems to have been rather marginal. One can write quite a satisfactory
history of the two services in the two countries with hardly any reference to
EU policymaking or regulation. Of course there have been some impacts
– police cooperation through Europol or the European coordination of
pharmaceutical regulation in healthcare – but at no time do these either
head the domestic policy agendas or dominate operational management
at the local level. The domestic effects of EU membership have, as many
writers have observed, varied enormously according to, inter alia, the com-
petencies set out in the treaties (for example Scharpf, 1999). We happen to
have chosen two sectors where the competencies of the EU institutions are
at present very limited. This may not continue indefinitely – for example,
as we write (2009), the Commission has produced a draft regulation con-
cerning the right of EU citizens to access health services in countries other
than their own – but up to now it would be hard to argue that the EU was
a key actor in either hospital policy or police policy.
Next, to reflect more systematically on our findings, we should return to
the main questions raised in Chapter 1, namely:
One striking difference between the two countries lies in the pattern and
processes of elite networks. In Belgium these networks span the federal/
regional and local levels, and to a considerable extent they are also cross-
sectoral. It is not merely that the mayors know the ministers, it is that the
mayors often are ex-ministers, and vice versa. This perhaps creates the poten-
tial for a kind of informal coordination which is more difficult in England:
Before the direct elections of the regional parliaments (1995) all members of the
community and regional parliaments . . . were also members of the Chamber or
the Senate. One week they would sit as a member of the federal Parliament, the
other week as a regional/community MP. Through this personal combination of
mandates, co-ordination between the parliamentary groups . . . at the federal and
sub-federal levels did not pose a major problem. (De Winter et al., 2006, p. 941)
In England the links are more distant and, of course, the greater size of
the country means that the elites themselves are bigger and more diversi-
fied. On the other hand England does not experience the strong division by
language and community that characterizes the Belgian situation. It may
be that the Flemish elite finds it easier to coordinate multilevel governmen-
tal policymaking than is usually feasible in England, but for the Flemish
elite to coordinate with the Walloon elite is another story altogether (De
Winter et al., 2006, p. 938).
These differences in elites and the cleavages between elites cast the whole
issue of autonomy in a different light. Increasingly, during the period under
study, the question in Belgium has become less what autonomy central
government will allow to the regions and municipalities, as what federal
actions the regions are prepared to put up with. The English assumption
of a powerful centre, dictating the terms of multigovernment interactions,
is wholly inappropriate across 50 km of water. Here we confirm something
that has been noticed by other writers. Klijn, in a general piece on net-
works and governance observes that:
[O]ne could argue that the growth of governance networks in the UK differs from
that seen in many other countries in that it has a more strongly instrumental/
managerial and vertical flavour. (Klijn, 2008, p. 515)
● History matters.
● Institutions matter.
● Developments over time often exhibit certain patterns.
Here we will revisit that introductory template. First, we will look again at
the list of what we thought policy was not, a list drawn from the general
public policy literature. We need to check how far our Anglo-Belgian
investigations confirm or refute those starting assumptions. Then, second,
we will discuss what patterns are discernable in our hospital and police
stories.
Our first ‘policy is not’ statement in Chapter 1 was that policy was not
usually ‘an instrumental-rational business, with clear goals, objectives,
calculations of costs and benefit and so on’. The police and hospital stories
show that policies could usually be dressed up in this rational garb, but
that in practice there were also usually contradictions, ambiguities, con-
flicts and accidents as well. Take ‘community policing’. It was certainly
instrumental and rational in the sense that politicians and police were
responding to public concerns and fears (as amplified by the media). Yet at
the same time it was a slippery concept, referring to quite a range of groups
and types of location or administrative unit, and to a pot pourri of tech-
niques and approaches. ‘It is hard to determine what precisely is entailed
in community policing or what precisely constitutes the community focus’
(Tilley, 2008, p. 40). Furthermore, governments have generally avoided
giving explicit guidance on how this top priority is to be balanced against
the other ‘top priorities’ such as terrorism or organized crime. Finally,
insofar as a good deal of research showed that routine patrolling of the
streets was a pretty inefficient way of using expensive professionals, it was
irrational to put more effort into it (Audit Commission, 1996).
Community policing also illustrates our second ‘is not’: that policy
usually is not a neatly staged process leading from formulation to imple-
mentation to evaluation. Community policing seems more like a hardy
perennial – something that sprouts up suddenly (especially after some race
riot or community collapse) and is then formulated and elaborated after
its introduction as a headline concept. Subsequently it gradually fades
from view until the next prominent local disturbance, at which point it
is resuscitated and dressed up again. Evaluations are occasional and fre-
quently inconclusive. Nevertheless, it is very hard to be against community
policing – it is widely accepted as a ‘good thing’.
Our third caveat was that what finally gets done (or left undone) often
Reflections on theories of change 141
does not closely match the policy as originally announced. This is certainly
not a universal divergence, but it is a frequent one. When we look at the
history of the KU Leuven (KUL) hospital we see its founders working
with a vision of a particular kind of integrated hospital that would be the
leading institution of its kind in Flanders and Belgium. Remarkably, 35
years later, that vision has been broadly realized. Elsewhere, however,
divergence is common. The huge post-Scarman police effort in England
during the 1980s to improve police–community relations, especially in
ethnically mixed areas, did not prevent later racially fuelled urban dis-
orders and neither did it enable the police, in the twenty-first century,
to reach their equal opportunity recruitment targets, or to avoid well-
publicized tensions within the Metropolitan Police and other forces. To
put it generously, cultural change is very slow. In Belgium, we can refer to
the government’s efforts to moderate the growth of the hospital budgets
by increasing the fraction of the cost of hospital visits that patients would
have to pay themselves. Instead of causing patients to refrain from seeking
unnecessary hospital care, initially it mainly led to a rapid growth in
top-up insurance to bridge the gap between what the basic mutuality
insurance would pay and what the hospital charged. In both countries
we have seen periodic initiatives and policy pronouncements intended to
eliminate police corruption but it is clear that, though they may ameliorate
the problem, such corruption is actually endemic, always able to take new
forms, and is particularly associated with certain roles or tasks (Punch,
2008, pp. 51–3).
The idea that policy simply flows from the head of one great leader
seldom finds any purchase in our histories. Even if the Leuven hospital
grew from the vision of a small number of medical academics, its realiza-
tion crucially depended on a wider network of Flemish politicians. Even
if Chief Constable Ken Jones vigorously led the Sussex force towards a
tighter, more performance-driven approach to policing, he did so with
internal and external allies, and (not least) in the knowledge that his
approach commanded enthusiastic support from ministers and the Home
Office.
Similarly, the idea that policymaking begins at the top and flows down
organizational hierarchies does not fit with several important elements in
our tale. The new Leuven hospital was born of local enthusiasts, although
enthusiasts who were wise enough to do their international homework
and evaluate a wide range of hospital design concepts. They went on to
persuade the upper reaches of the university and then the Flemish and
Belgian political hierarchies. Even in the far more centralized English hos-
pital system there are many instances of local innovations or pilots which
attract attention and then get ‘rolled out’ to other locations. The Royal
142 Continuity and change in public policy and management
Having dealt with this list of mis-assumptions – and confirmed that as far
as our study is concerned they are such – we can now turn to the question
of patterns of change. To provide us with some framework for this discus-
sion we repeat the BEST diagram originally introduced in Chapter 1 – see
Table 8.1
Before we dive into the detail of this analysis it is necessary to pause
and remind ourselves of the levels at which we have been operating. In
Chapters 3 and 4 we described sectoral policies at national levels. In
Chapters 6 and 7 we described local policies in an English city and county
and a Belgian city and its hinterland. The book has therefore been pitched
mainly at what, in policy studies terms, is the middle or meso-level, with
a few more detailed, micro-level touches to fill out the local narratives.
Other than very briefly in Chapter 2 we have not been much engaged at
the macro-level. We have not traced, for example, the changing shape
of the welfare state in either country, or in explained overall shifts in the
national machineries of government. This has implications for the kinds
of theoretical apparatuses which are most appropriate for our enquiry. In
short, it has pushed us towards the idiographic end of the nomothetic–
idiographic spectrum that was mentioned in Chapter 1, section 1.10. Kay
puts the issue like this:
Result of change
Within path/incremental Radical/transformation
Process of Gradual A. Classic B. Gradual, but eventually
change incrementalism fundamental change
TORTOISE STALACTITE
Abrupt C. ‘Radical D. Sudden, radical
conservatism’ – rapid change
return to previous ways EARTHQUAKE
BOOMERANG (punctuation)
Thus our study is of a more ‘granular’, contexed kind, and the patterns
we may (or may not) find are unlikely to fit the formal and usually highly
abstract nature of broad, hypothetico-deductive generalizations. (We will
return to this issue in greater depth when we look at doctrines of compari-
son in the final chapter.)
To come now to the substance, the first, and most important point, is
that our Anglo-Belgian comparison has yielded examples of all four types
of change, often mutually interwoven. To develop this point, change does
not seem to come in one package. More typically, several different types
of change may be going on in a given sector during a given period, each
layered upon and tangled with the others. This complexity appears to
have been true for both sectors and both countries – it thus transcends
the important institutional and cultural differences between Belgium and
England that we have previously commented upon. Our two-country,
two-sector comparison therefore gives no grounds for believing that a par-
ticular kind of political system produces one these patterns of change (say
Tortoise) while another system produces another (say Earthquake). Both
systems produce a mixture of the different patterns indicated in Table 8.1
However, there is a different frequency to the occurrence of change,
as between the four categories. On the whole it is the top line – gradual
change – which seems most common. Sudden, unreversed, radical change
(Earthquake) is the rarest, and we have few examples of it in our study.
The most obvious one is the creation of the new university hospital at
Leuven, but even that is not a 100 per cent clear case. In its favour one
might say that an entirely new, large facility, of an innovatory design,
was swiftly built and, when put into operation, employed what were (for
Belgium) novel systems of human resource management, and a ‘progres-
sive’ philosophy of integrated, multidisciplinary care. This is by any
yardstick a major achievement, but its position as a pure Earthquake is
slightly diluted by two further observations. First, for large hospitals,
the move from vision to working reality cannot be particularly fast. The
‘big idea’ for the new KU Leuven Ziekenhuis dates from the late 1960s
and early 1970s, but the inauguration of the hospital did not take place
until 1985. Whether one can describe this as ‘sudden’ is doubtful. Second,
the ideas involved, while new for the Belgian hospital scene, were largely
(consciously, systematically) taken from the existing, international profes-
sional discourse of the time.
Reflections on theories of change 145
Since the first chapter we have made use of two broad categorizations of
policy change – the three levels of Hall, and the fourfold categorization
into Boomerang, Earthquake, Stalactite and Tortoise (BEST) which we
have ourselves built out of an original schema suggested by Streeck and
Thelen. The Hall scheme focuses mainly on the level of the change – small,
medium or big, as defined by whether only settings, or instruments or
paradigms themselves are altered. Our BEST schema focuses on both the
level of change and its modality: has the process been brutal and short
(Earthquake) or slow but directional (Stalactite) or meandering and
gradual (Tortoise) or fast but short-lived (Boomerang)? The two schemes
do not map directly onto each other, except that Hall’s level three change
would clearly have to be the product of either Earthquake or Stalactite.
It has been clear from the beginning that assigning actual, real-world poli-
cies to cells within either of these typologies entails a good deal of judgement
on the part of the analyst. Neither scheme fits neatly over all the varieties of
specific policies in a messy world. Policies cannot sensibly be displayed like
stamps in a stamp album, neatly lined up in rows according to indisputable
criteria of date, origin, denomination and so on. There is a question, there-
fore, of how far we can push such typologies. Whilst they seem tolerably
clear and useful at a macro-level or over a long time span, how far can we
take them into our meso-world before they lose their discriminatory power
and become factors that complicate rather than simplify?
Reflections on theories of change 147
Despite much discussion, we are entirely not sure ourselves what the
correct answer to this question should be. In the remainder of this sub-
section we therefore make a highly tentative attempt to apply the BEST
scheme and the Hall three-level model at a level of greater detail. We will
do this at the local rather than the national level, partly because these
types of scheme have been less often applied there, so the exercise is more
original, and partly because we anticipate that going ‘down’ to local level
may provide a stiffer test for these classifications.
First, therefore, we have to go back to the ‘local’ chapters – Chapters
6 and 7 – to select key changes for analysis. This is in itself problematic,
insofar as we cannot regard our selections as representative of the total
population of policies in any statistical sense. This book has never claimed
to be able to describe all policies (indeed, it is doubtful if any book could).
Thus there are whole areas of hospital policy and police policy which have
fallen entirely outside our research. We have not looked at police recruit-
ment policy, for example, nor at police financial management. Neither
have we examined hospital safety policies, or the NHS initiatives on junior
doctors’ working hours – important though all these issues were.
What we did do at local level was to ask which changes had been impor-
tant, and which things had stayed broadly the same (see Appendix). So
if we go back to Chapters 6 and 7 and ask what the important changes
seemed to have been, we come up with a list like that in Table 8.2. The
capital letter in brackets after each entry signifies our best assessment as to
whether the change was B, E, S or T.
How well do these fit into Hall’s model and BEST? We can take them
one by one.
The double failure to secure a new greenfield site acute hospital for
Brighton (mid-1970s and late 1980s) is hard to characterize using BEST.
That is because, first, it was a non-change rather than a change – the defeat
of local policymaking rather than an achievement. Looked at from a
macro-perspective, however, the two rejections were rather different deci-
sions, one from the other. The first was the result of an Earthquake on a
national, indeed global, scale. Following the oil price shocks of the early
1970s the British economy wobbled badly, and the government of the day
felt itself bound to seek International Monetary Fund (IMF) aid – a night-
mare shift of paradigm. One price of that aid was a set of deep cuts to the
whole range of public spending programmes. The new hospital at Falmer
was just one local example of many casualties in these fiscally fraught
times. The second rejection was a result of a more incremental policy. It
was the national policy, explicitly pursued from the mid-1970s onwards
for many years, to ration NHS new investment at the margin in favour
of those parts of the country which suffered the highest morbidity and
148 Continuity and change in public policy and management
mortality. This was a perfectly rational policy in its own terms, and incre-
mental when viewed from the Olympian perspective of the Department
of Health. From a Brighton point of view, however, it meant that it was
always going to be hard to make a strong enough case for a single, big
new investment. This policy might be thought of as a Tortoise, because
it involved a very slow, step-by-step migration towards equalling out the
historically unequal distribution of resources within the NHS. (Had it ever
reached its goal it might have been recognized as a Stalactite at national
level, but that is not an issue we can deal with here.) Yet for Brighton (and
other localities) it had a sharp, singular result – not half a new hospital
or a quarter of a new hospital, but no new hospital at all. Overall, then,
the BEST framework is not particularly helpful here, and neither is Hall’s
three-level model.
The long-term centralization and reprovision of acute facilities on the
Royal Sussex County Hospital (RSCH) site was a Tortoise type of change.
It went on, bit by bit, for virtually the whole of our 40-year period. It
accelerated in the 1990s, when it had become clear that the alternative
of a greenfield site was as good as dead, and when the Private Finance
Initiative made the availability of capital somewhat less constrained. In
the period 1998–2008 the amount of new investment on the RSCH site
has been very considerable, but it would be incorrect to represent this as a
deliberate long-term policy all along.
Reflections on theories of change 149
clear what effects this might have had on actual police operations, and still
less on outcomes in terms of public order. It has waxed and waned on local
agendas, but had basically grown from the late 1980s onwards and had
later become a part of the new national orthodoxy, so that Brighton was
one of the police authorities which could claim to be ahead of the field. We
therefore classified it as a Tortoise, although we suspect that some of our
interviewees (but not others) would have said it was more of a Stalactite –
something which accumulated until it was a striking feature of the police
landscape. In Hall’s terms it is again hard to place. Some might say it
was (for England) such a new idea that it should be seen as a paradigm-
breaking third-level innovation. Others would argue that it was merely
going back to earlier days – righting the balance a bit after the techno-cop
period of the 1970s and early 1980s. Certainly it entailed the creation of
some new policy instruments (level two).
The growth of performance indicators (PIs) was another much-men-
tioned change, and one that we would see as a Stalactite. When PIs were
first used in the late 1980s they did not seem particularly central to police
operations – an interesting new managerial gimmick, perhaps, but not
much more. Gradually, however, the framework of PIs grew in scope and
sophistication, and the government started to use them more and more for
assessing the competence of police forces. By the end of our period they
had almost become a way of life. In Hall’s terms they are clearly level two
– a (once) new policy instrument.
Performance indicators were one element in a wider phenomenon –
growing Whitehall intervention in the management of police forces. This
movement also involved, inter alia, the reshaping of the police authorities in
the mid-1990s and the introduction of regular National Policing Plans from
2002. We would classify this as a Stalactite because it involved many differ-
ent moves over a decade or more in a consistent direction (including some
that were diluted or defeated by the police themselves). No single move was
in itself an Earthquake, but they accumulated. This process reflected a con-
scious and deliberate intent that was, to a large degree, shared by the Major
Conservative government and its New Labour successor. In terms of Hall’s
model this succession of measures certainly represented level two change
(new instruments) and there is a case for saying that if one compared, say,
2005 with 1975, there had actually been a shift in the paradigm governing
relations between the government and its police forces.
The 2001 catharsis of Paul Whitehouse’s resignation is another develop-
ment that is hard to classify. Whilst virtually all our interviewees thought
it had been important and traumatic, the deeper question is whether it
made a big difference to the way the Sussex police were organized. We are
inclined to answer ‘Some but not much.’ Certainly there was a change in
Reflections on theories of change 151
mood and style under Ken Jones, and the use of perfomance indicators
came to the fore, but these strike us as changes of emphasis rather than sig-
nificant changes of policy. The main lines of national government policy
continued to be applied within the Sussex force. So it does not qualify as
an Earthquake, but neither was it a Boomerang, Stalactite or Tortoise. It
was a dramatic event but not one that can be readily fitted into either the
Hall model or BEST.
Finally, the long-term, multifaceted processes by which the Sussex
Police (and other English forces) became more specialized and profes-
sional are probably our best example of a Stalactite. Sometimes this direc-
tion was facilitated by central policies (PACE led to a huge training effort
for nearly all police) and sometimes it was more a matter of local initiative
(setting up a special unit for this or that; developing special techniques for
dealing with ‘problem estates’). Taken as a whole this process would easily
qualify as a Hall level two change and it could be argued that, over three
decades, it has almost amounted to a level three shift of paradigm (from a
practical, generalist force to a high-tech, specialized force).
The first key change for the Leuven police was the 1976–77 restructur-
ing. This is another issue which is hard to classify. In one way it seems
of great consequence – an Earthquake – but in another it appears to be
of much smaller importance: just a change to organizational structures
which, although big in terms of numbers of forces, had little immediate
effect on police cultures or operational practices. Some of our interviewees
took this latter view, some the former. We cannot, with any confidence,
find a box to fit it.
The Pinksterplan of 1990 is also less than straightforward. We tend
to see it as an early and important step in a long process that eventually,
with the ‘help’ of external events, led to the major structural reforms of
1998–2001. We would therefore bracket this with the Dutroux drama
of 1996–98 and the 1998 Police Reform Act. They form a set of factors
which eventually produced an Earthquake, but only after a Stalactite-like
process of negotiations and attempted reforms that stretched back for a
decade. Dutroux was an Earthquake that provided the window of oppor-
tunity that, in some sense, the reformers had been waiting for. It enabled
them to cement their intentions in the new law, which itself became an
Earthquake for structural and professional reforms.
Last of all we should refer to the processes of professionalization and
specialization which affected the Leuven Police as they did the Sussex
Police. These fit quite well to the notion of a Stalactite. As in Sussex, there
were many sources, central and local (and occasionally even international)
but they tended to flow in the same general direction – towards quality,
standards, training, professionalism. It is a process which still continues.
152 Continuity and change in public policy and management
According to Hall’s model, it mainly involves level two changes, but there
is a judgement to be made about when so many of these have accumulated
that the new ensemble can be regarded as a new paradigm.
To sum up this subsection, it seems that neither the BEST nor Hall’s
three-level scheme work very easily when applied to detailed policy nar-
ratives at meso- and lower levels. Most of the key developments which
we identified can, with a bit of intellectual effort, be pushed into one of
the BEST boxes. Sometimes they fit quite snugly. But in quite a few cases
there is some uncertainty about which box is most appropriate, and in one
or two it is clear that none of the four boxes fits very comfortably at all.
One obvious untidiness about real world changes is that they may change
speed over time. For a few years they may Tortoise along, and then speed
up during a particular period and subsequently slow down again. To take
this point further, the BEST scheme (and its Streeck and Thelen forebear)
really only offer a two-speed gearbox – gradual and abrupt. But there is no
reason why policies should not proceed like a car in traffic, crawling here,
speeding up a bit when the traffic eases, and occasionally, on the open
road, travelling really fast. In short the patterns of change are more varied
than a simple two-by-two classification can fully capture. We need at least
a four- or five-speed gearbox, and the capacity to make gear changes as the
conditions in the policy traffic and the wider environment change.
The principal value of the two schemes (and it is considerable) is there-
fore heuristic – they do help to organize broad ideas and generate useful
questions. Trying to use them obliges the user to think hard about what
the shape of the policy was, and how best to describe apparent changes
of pace. (And they do fit quite a number of episodes, even if they do not
fit all.) The performance of BEST (or Hall’s three levels) is less impres-
sive when they are deployed as empirical sieves for ordering detailed data
and/or generating matrices of policy performance. The problems in this
latter role are just too many. One would need a population of all poli-
cies in order to establish true representativeness, and that can seldom be
achieved. Policies would need to be: (1) readily identifiable; (2) mutually
exclusive; and (3) roughly comparable units, so that they could be mean-
ingfully counted. But in fact policies overlap and interact, and different
policies have vastly different scopes and weights and lifetimes. And there
is the further complication that a T at national level may constitute an E at
local level, or a local B may not register as anything important at national
level at all.
Thus to apply BEST or Hall’s model as the basis for a comprehensive
quantitative analysis would be highly problematic. It would be counting
together not so much apples and pears as branches, trees and forests.
Furthermore the Hall scheme has an additional limitation, from our point
Reflections on theories of change 153
authorities ever to win a big enough slice of the tightly controlled NHS
capital budget was the single most important reason why no new hos-
pital appeared in the area between 1965 and 2005. This inability can be
explained in terms of the particular resource allocation rules in place at the
relevant moments when bids were made, and more generally in terms of
the existence of a national system that insisted on elaborate comparisons
of all bids from right across each region and, indeed, the whole country.
The Leuven hospital never had to go through such a nationally orches-
trated budgetary trial-by-comparison.
Organizational: organizational routines can act as powerful mecha-
nisms for continuity. Sometimes these routines may actually be embedded
in law, which further strengthens their formative power. But even routines
that are not prescribed by statute can constrain the rate and direction of
change. In interview senior staff at KU Leuven Ziekenhuis were convinced
that the medical payment system put in place by Dr Peers in the early 1970s
had helped to sustain organizational loyalty and coherence. ‘Once the new
salary system was accepted, it was a basic turning point’ (interview 5)
Cultural: cultural norms are usually self-reproducing and help give an
organization or group a sense of its own identity and history. Often they
act to reinforce continuity and stability but, in theory at least, it is per-
fectly possible that an organizational culture could exist that prizes and
rewards continuous innovation and change (as was rumoured to be the
case at Microsoft). In our study cultural norms were frequently noticeable,
and were sometimes emphasized by interviewees. There is, for example,
the professional culture of a university teaching hospital, which values
research, high skill and technical innovation. One could say that this
makes it ‘naturally’ an ever-changing, expanding, renewing kind of estab-
lishment, at least at the micro-level. But this culture also has wider conse-
quences. It means that the staff are constantly developing new treatments,
some of which run ahead of the categories in which third-party payers
reimburse the hospital for its services. Indeed, this was one component in
the financial crisis at KU Leuven Ziekenhuis in the mid-1990s. In the case
of the police several of our English interviewees stressed that the ‘street
culture’ of the lower ranks had largely survived the many organizational
and technological changes through which the service had gone. They
characterized this culture as one which valued pragmatism and a ‘can do’
attitude and emphasized team loyalties, but combined this with deep scep-
ticism about politicians and ‘bosses’, plus a fairly jaundiced view of human
nature more generally. Most of all, policing was a career, not just a job.
Policy actors are not to be thought of as the passive victims of this array of
mechanisms. Some mechanisms may operate ‘behind the backs’ of some key
decision-makers, but for the most part the latter are necessarily aware of the
Reflections on theories of change 157
It is important to note that for a particular policy paradigm to fade and, hence,
for institutional change to occur, it is not enough that a single agent comes to
recover into consciousness and stop behaving in accordance with it. Rather, the
population of agents to which the rule applies must, collectively, stop behaving
in accordance with the rule. (Kay, 2006, p. 71)
Finally, it should be clearly apparent by now that the search for different
kinds of mechanism requires more than one set of methods. Researching
long-term cultural change cannot be satisfactorily accomplished using the
same tools as are deployed for researching the immediate antecedents of
Earthquakes. The role of organizational standard operating procedures
cannot be best investigated with the methods we would use for exploring
shifts in, say, elite self-perceptions of Flemish autonomy. If we conclude
that many types of processes are at work in promoting or inhibiting
change, then we must accept the logical consequence that many kinds of
research project will be needed to explore these different influences.
8.5.1 Introduction
This is not just a comparative study of continuity and change in two public
services in neighbouring countries, important though we think that is in
itself. It also has a direct bearing on some major contemporary controver-
sies in the social sciences. In particular we believe it sheds some light on
Reflections on theories of change 159
survey similarly showed Belgians to be more satisfied with their health care
system, and also showed quite big differences (in favour of Belgium) of
public perceptions of how well run the health care system was (Mossialos,
1997). Such surveys of popular perceptions are not the firmest ground on
which to found public policymaking (perceptions may be heavily context-
dependent and even volatile), but we have little else to go on.
We now turn to the point about responsiveness. Lijphart focused on
responsiveness in the sense of the representation of a spectrum of views
in the legislature and, beyond that, the executive. We would not for a
moment deny that that is a core issue for any democracy, and on that score
the English two-and-a-bit party system and single-party cabinets look
fairly insensitive to variety. There is, however, another kind of responsive-
ness, which also merits attention. That is the day-to-day responsiveness
of public services to the immediate and longer-term needs of individual
citizens. This ‘street-level responsiveness’ may go hand-in-hand with
traditional party-political representativeness, but there again, it may not.
Highly differentiated political responsiveness to voters in terms of elected
representatives, party programmes and finely tuned rhetoric may coexist
with unresponsive service bureaucracies and/or unresponsive public service
professions. It may also coexist with patronage and partiality in the staff-
ing or provision of public services – a lack of impartiality. In these respects
consensus Belgium does not seem to perform so well (Brans et al., 2006;
De Winter and Dumont, 2006)
Here we refer to the seminal work of the leading German scholar Fritz
W. Scharpf. In a series of publications dealing with both the institutions
of the European Union and national governments Scharpf has drawn a
distinction between ‘two faces of democratic self-determination’. These
he terms ‘input-oriented legitimization’ and ‘output-oriented legitimiza-
tion’ (Scharpf, 1999, Chapter 1). The former refers to a way of thinking in
which political choices are legitimate to the extent that they reflect the ‘will
of the people’ – that is, they can be derived from the authentic preferences
of the members of a community. The latter sees the legitimacy of choices as
depending on results – on whether they ‘effectively promote the common
welfare of the constituency in question’ (Scharpf, 1999, p. 6). This connects
with the nature of the political system because it is argued that systems of
proportional representation tend to deliver higher levels of input legiti-
macy, because they promote a more differentiated and variegated pattern
of parties, more accurately reflecting the preferences of the wide range
of groups that go to make up a modern, ‘post-industrial’ society. On the
162 Continuity and change in public policy and management
other hand majoritarian systems may find it easier to choose and imple-
ment coherent, focused policies and programmes that will enhance output
legitimacy. Scharpf himself argues that: ‘input-oriented authenticity and
output-oriented effectiveness are equally essential elements of democratic
self-determination’ (Scharpf, 1997, p. 19).
These contrasts are, of course, quite close to – although not identical
with – those previously discussed in connection with Lijphart’s views on
the relative performance records of consensual and majoritarian democ-
racies. Scharpf does not insist, as Lijphart does, that consensual regimes
have higher ‘input legitimacy’, but it is not hard to read that message into
his argument (see, for example, Scharpf, 1999, p. 14, footnote). Some of
the same qualifications that we made about Lijphart can therefore be
reiterated. There is a big gap between headline political choices (‘the gov-
ernment decides on more competition in health care’ or ‘ministers go for
community policing’) and the experience of citizens on the ground. Both
the big decisions and the local experiences may contribute in some general
way to ‘output-oriented legitimization’ (or the lack of it) but if there is an
obvious disconnect between the two – as there may be – we may expect
a growing disillusionment with the political elite. Just such a disconnect
– aided and amplified by more and more aggressive mass media – is said
by many commentators to have begun in both countries, indeed interna-
tionally. However, this is a tricky subject: careful examination of such
international survey data as are available suggests that fluctuations in trust
levels cannot accurately be depicted as a general decline (Van de Walle et
al., 2008)
Scharpf himself is centrally concerned with the differences in patterns
of input and output legitimacy at the interacting levels of national gov-
ernments and EU institutions. He is preoccupied with the problem that
a competitive market at the European level may undermine the ability of
individual member states to sustain relatively generous welfare states.
Here, however, we have been concerned with a type of difference which
his analysis does little to address: that between national-level policymak-
ing and local policy implementation. In effect this complicates his model
because it adds a third layer – and one of great importance to most citizens
on a day-to-day basis – at which both input and especially output legiti-
macy is created or destroyed.
Relevant and reliable empirical evidence is hard to come by on this
issue. Eurobarometer does have a series of surveys stretching back to the
1970s in which citizens are asked: ‘On the whole, are you very satisfied,
fairly satisfied or not satisfied at all with the way democracy works in your
country?’ Extracting from this series (and reiterating the earlier caveats
concerning surveys of popular perceptions) we get Figures 8.1 and 8.2.
Reflections on theories of change 163
70
60
50
Belgians very satisfied
40 UK very satisfied
%
30 UK fairly satisfied
Belgians fairly satisfied
20
10
0
1976 1982 1988 1994 2000 2006
Notes: ‘Don’t knows’ are not shown to avoid crowding the graph.
Figure 8.1 How satisfied are you with the way democracy works in your
country?
45
40
35
Belgians not satisfied
30 at all
25 UK not satisfied at all
%
20 Belgians not very
satisfied
15
UK not very satisfied
10
5
0
1976 1982 1988 1994 2000 2006
Notes: ‘Don’t knows’ are not shown to avoid crowding the graph.
Figure 8.2 How satisfied are you with the way democracy works in your
country?
164 Continuity and change in public policy and management
Notes:
1. The third column expresses expenditure in real terms, adjusted by a GDP deflator at
market prices, where 1949 – the first year of the NHS – is set at 100.
2 The NHS represents (2003) about 85% of total UK health expenditure.
Source: Harrison and McDonald (2008), pp. 11–12. originally from Office of Health
Economics.
Growth has been particularly fast since 2000, when the Labour govern-
ment announced its intention of raising UK health expenditure as a per-
centage of GDP ‘to the European average’. Hospital spending has more
Reflections on theories of change 165
than taken its share of the overall growth. If we look at change between
1997 and 2007 we find that the number of consultants in the English NHS
rose by 12 200 (4.6 per cent per annum) and of registrars by 18 850 (10
per cent per annum) (http://www.ic.nhs.uk/, accessed 11 November 2008).
NHS hospitals have been a ‘growth business’.
None of this means that the European single market will not push
national governments into welfare reductions in future, but it does mean,
so far as hospitals and health services are concerned, that it has not begun
to happen yet. It is possible, of course, that the impacts have for some
reason been concentrated in other parts of the welfare state – pensions,
unemployment insurance or social care – but it does seem that we can
say that Scharpf’s concerns about the impact of the single EU market on
welfare expenditure is not yet borne out by what is happening in health
care.
Next we look at evidence concerning citizens’ attitudes towards the
civil service. These data raise another question concerning the theories
of Scharpf and, more particularly, Lijphart. It is the question of where
the civil service (and the public service more generally) fit into their argu-
ments. That is not clear. On the one hand civil servants are not politicians
(although in many countries, including Belgium, senior civil servants are
appointed within a party patronage system). Therefore one could argue
that they play no role in either input or output legitimacy, and are not
part of the model. On the other hand the civil or public service is definitely
seen by many as part of ‘government’ or ‘the state’. Therefore it could be
argued that trust in the civil service would form one component in the
aggregate citizen judgement on the legitimacy of the regime. This is a point
of some importance because citizen perceptions of the civil service can and
do diverge from their perceptions of the political elite. In crude terms they
may trust their politicians but not their civil servants, or vice versa (or, of
course, they may also trust both, or neither).
Table 8.5 is extracted from the World Values Survey and shows the per-
centages of those surveyed who declare themselves as having a great deal
Source: World Values Survey and European Values Survey – see Van de Walle et al.,
(2008), p. 58.
166 Continuity and change in public policy and management
Note: The question to the respondent is whether s/he tends to trust the civil service or not.
The figures in the table are the percentage of respondents who tend to trust.
Source: Eurobarometer.
[C]itizens generally come into contact with the output side – with the adminis-
tration that is – far more frequently and intensively than they do with the input
side . . . public administration is the political system – as citizens concretely
encounter and experience it. The character of the administration is therefore
decisive for the way in which the political system is viewed. (Rothstein, 2008,
p. 18)
Rothstein and Teorell (2008) also stress the foundational role that
impartiality plays in generating citizen perceptions of regime legitimacy.
They argue that impartiality is even more important than effectiveness
or efficiency. This is an interesting argument but one that we have not
explored in any direct way in our study. Hypothetically, however, we
would expect it to lead to relatively lower ratings for Belgium, since the
Belgian system appears to exhibit more extensive and obvious party
political patronage than the English (in public sector appointments and
promotions, for example).
Finally, we turn to the Scharpf fear that welfare states will be under-
mined by the economically competitive logic of the EU single market.
Once again, our study of hospitals yields little evidence to support this
concern. Public expenditure on health care has actually grown in both
countries and, within that total, hospital spending is buoyant.
8.6 CONCLUSIONS
have each become more and more like themselves (rather than more like
each other). While specific ideas and techniques may have been picked up
in both countries (neighbourhood policing, patient-focused care and so
on) the ways in which these ideas and techniques have been introduced
have been very different. It is almost as if each country has moved closer
and closer to its own policymaking caricature – as if this cultural and
procedural element was the most immovable component of the whole
complex process of change. So Belgian policymaking in the early twenty-
first century was even more opaque, slow-moving, fragmented, bargained
and overtly politicized than two or three decades earlier, while English
policymaking was even more centralized, hierarchical and managerial.
If we are right about this, there must be a limit to it – a threshold at which
each style breaks down and either something else is put in its place or, at
the very least, there is a measure of incremental retreat. Furthermore – and
speculatively – it could be argued that, right at the end of our period, these
‘style thresholds’ were beginning to become visible. In Belgium 2007–08
saw a huge national political crisis during which the politicians could not
form a government and divisions between the various regions and parties
seemed to grow even deeper, with a kind of political paralysis setting in
and little progress being made on several crucial issues. Meanwhile in
England resistance to centrally dictated performance regimes and endless
centrally orchestrated reorganizations and initiatives grew louder and
louder. A Cabinet Office paper on public service reform was remarkably
apologetic about previous micro-management, Whitehall control and
‘giving schools, hospitals, local government and local communities more
power and responsibility to decide what is right for their area’ (Cabinet
Office, 2008, p. 42). Whether these fair words had any subsequent effect is,
of course, another question, but the whole tone of this paper spoke of a
consciousness that a particular style and approach had been pushed to its
sensible limit and beyond.
This idea of some kind of limits currently being approached is perhaps
a nice, speculative point at which to conclude our analysis. Alongside
the sense of limits, we should also close with the observation that – for
Belgium and England – the now classic academic model of policy stability
punctuated by rare episodes of radical change just does not fit the facts.
On the contrary, change of one kind or another has been constant in both
sectors and in both countries. It has been multilayered and untidy, but it
has happened. Even in the stereotypically slow-moving Belgium we find
big changes in the political system (party fragmentation, decentralization
of policy competences) right alongside sometimes agonizingly slow shifts
in implemented policies. And at the same time, in hyperactive England,
we notice considerable continuities, such as the impartiality of the civil
Reflections on theories of change 169
service, the tradition of operational autonomy for the police or the seem-
ingly inexorable growth of the acute health care budget. And last, but not
least, some very important changes accumulate gradually, without figuring
much on the explicit policy agenda at all. Our long-term view has helped
us to see that change takes a multiplicity of forms and has a multiplicity
of sources, endogenous and external, of which policymakers are only one.
Policy implementation – often ignored or relegated to a minor role by
political scientists and legal scholars alike – turns out to be very important.
This is no news to students of public administration, or to the millions who
use public services, but it is nonetheless a suitable point with which to end
the chapter. Formal government policy pronouncements may be the end
of complex processes of information gathering and political bargaining,
but they are only the beginning of the equally complex business of turning
fair words into feasible actions. Studies of change need to pay as much
attention to implementation as to formulation.
9. Reflections on doctrines of
comparison
9.1 INTRODUCTION
The previous chapter dealt with the substance of our conclusions and
further thoughts about our two public services in our two Western
European countries. This final chapter deals not with the substance but
with the ways in which that substance has been collected, assembled
and presented. We have told a pair of intertwined stories (one local, one
national) for each country, and we have then compared these pairs, one
with another, looking for similarities, differences and general patterns over
time. Each has been a complicated narrative, with many different events
(for example riots, decisions, changes of government) and influences (for
example finance, technology, culture). Each has offered ‘thick description’
but has also engaged with a range of theoretical issues (the effects of speci-
fied types of political and organizational regimes; path-dependency; causal
mechanisms; and so on). It is therefore time to ask what kinds of stories
these have been, how persuasive and reliable they are, and what kind of
explanations they will bear.
This is, in effect, a methodology chapter, but we have not called it that
but have chosen instead to title it ‘Reflections on doctrines of comparison’.
That choice is intended to send two important messages. First, the use of
a plural (doctrines) emphasizes that there is no one best way to undertake
all international comparisons – no single, stratospheric, gold standard
against which all other approaches can be graded. Second, use of the term
‘doctrines’ suggests that the selection of a methodology by a researcher is
not an exclusively scientific activity. Of course it is a scientific and technical
decision, but not only a scientific and technical decision. Epistemological
affiliations also frequently have ideological, political or even ethical
dimensions to them (Bevir and Rhodes, 2006, pp. 52–5; Hawkesworth,
2005; Lynn et al., 2008; Phillips, 2000, Chapter 11). Both these messages
are somewhat controversial (not just recently, but rather perennially so).
Nevertheless, we believe that many comparative scholars would more or
less share the position that we are about to expound, although certainly by
no means all of them would.
170
Reflections on doctrines of comparison 171
Some of those colleagues who would disagree with our position would
insist that there is a clear hierarchy of approaches to knowledge acquisi-
tion, with something like a scientific laboratory experiment standing at
the top, statistical testing of clearly formulated hypotheses against large
populations a little further down, and descriptive case studies like ours
somewhere near the bottom (Hamm, 1988 offers a particularly clear and
well-argued example of this, although it was focused on clinical decision-
making rather than policymaking). This hierarchy is still widely taught
and practised, especially in the USA, and in countries where US social
science is a dominant intellectual import. In subsequent sections of this
chapter we characterize it as ‘scientific orthodoxy’, or the ‘variables
paradigm’. It is also frequently called ‘modernism’, especially by its oppo-
nents (for example Adcock, 2005; Bevir and Rhodes, 2006). Even if true
unreconstructed orthodox positivists are nowadays becoming harder and
harder to find, the shadow of their project for an objective science still falls
heavily over many labouring social scientists – across a variety of theoreti-
cal persuasions.
Since at least the late 1980s, however, others opposed to this paradigm
have claimed that there has been a turning
away from, if not against, the idea of a social scientific practice derived from
a model of human behavior abstracted from the physical and/or natural sci-
ences, denuded of the human traits of researchers and researched; and towards
a rehumanized, contextualized set of practices. (Yanow and Schwartz-Shea,
2005, p. xii)
We begin with what is probably still the most orthodox, mainstream posi-
tion in many university settings and in many parts of the world. We will
first describe the central claims of this doctrine and then critique it. In a
brief treatment we cannot examine all the subtle variations on this posi-
tion, but we will try to do it justice by basing most of what we say on one
of the most brilliant and widely used texts produced within this approach,
King, Keohane and Verba’s Designing Social Inquiry (1994). For conven-
ience we will refer to this as KKV.
KKV insist that there should be one unified inferential logic that unites
all real social science (p. 6). They argue that this logic is often at its clearest
in quantitative work, but that it should also apply to qualitative work. It
should apply equally to description (‘using observations about the world
to learn about other unobserved facts’) and causal inference (‘learning
about causal effects from the data observed’ – p. 8). They see all the social
sciences as being centrally involved in producing generalized, simplified
descriptions and explanations. Further, ‘our capacity to simplify depends
on whether we can specify outcomes and explanatory variables in a coher-
ent way’ (p. 10). Note here the emphasis on what can be observed as the
basic ‘feedstuff’ of both description and explanation. Hence the advice ‘to
Reflections on doctrines of comparison 173
The language of variables has become almost second nature to many social
scientists, including some of those who would not accept other aspects of
the KKV approach. Yet it is also possible to view ‘variables thinking’ as
narrow, mechanical and unrealistic. The eminent sociologist, Andrew
Abbott, in an article that looked back at the Chicago School of sociolo-
gists, argued that the location of social facts in time and space was crucial
to understanding. Yet the dominant ‘variables paradigm’ often promptly
removes them from these locations. He put the matter forcefully:
[N]ot only do variables not exist in reality, they are misleading even as a
nominalist convention. For the idea of a variable is the idea of a scale that
has the same causal meaning whatever its context: the idea, for example, that
‘education’ can have ‘an effect’ on ‘occupation’ irrespective of the other quali-
ties of an individual, whether those qualities be other past experiences, other
personal characteristics, or friends, acquaintances, and connections . . . The
Chicago view was that the concept of net effect was social scientific nonsense.
Nothing that ever occurs in the social world occurs ‘net of other variables’. All
social facts are located in contexts. So why bother to pretend that they are not?
(Abbott, 1997, p. 1152)
Both these show quite strong correlations with the status of being a
consensus democracy, so Lijphart takes his hypothesis as demonstrated.
From a different theoretical and methodological perspective, however,
we can see problems with this way of reasoning. These ‘variables’ are
each hugely complex. It is not just that the value of the variable is going
up and down, it is that the substance of the variable itself is somewhat
plastic. On the one side, consensus democracies (and majoritarian democ-
racies for that matter) are very different among themselves. Consider
the many differences between, say, hyper-stable yet modernizing polities
like Sweden and Denmark and a fragmented, traditional system such
as Belgium. On the other hand ‘decommodification’, whilst certainly
an intriguing concept, is both abstract and complex – and very difficult
to measure. The original attempt to quantify it by Esping-Andersen
(1990) turned out to be highly controversial, and was strongly criti-
cized (for example in Castles and Mitchell, 1993). Furthermore ‘social
expenditure’, using the Organisation for Economic Co-operation and
Development (OECD) definition, is also a tremendous rag-bag of differ-
ent programmes in each country and, as the OECD itself acknowledges,
is not measured in exactly the same way by each member state contrib-
uting to the statistics. So ‘decommodification’ and ‘social expenditure
as a percentage of GDP’ are hardly thermostat-type measures, showing
(say) 18 °C in majoritarian states and 23 °C in consensus states. Each
political system, each element of ‘decommodification’ and each budget
line of social expenditure is, as Abbot suggests, located in a particular
historical, political and institutional context. To take them away from
these contexts, aggregate them and then measure the correlations between
Reflections on doctrines of comparison 175
them is, indeed, a great leap of analysis. It gives us a number, but what
does that number mean? It is not a number like temperature, or the
number of votes, or the amount of money in your bank account. It is a
number measuring academic abstractions of academic abstractions, and
in consequence its precise import is hard to locate. Furthermore, the
measures themselves tell us nothing about whether the citizens of consen-
sus democracies actually believe that their governments are kinder and
gentler (or, indeed, whether the citizens of the majoritarian democracies
see their governors as less caring). Still less do they tell us why citizens
believe whatever they believe. And even if the hypothesis is true (consen-
sus democracies are kinder and gentler, on average), by themselves the
correlations made by Lijphart give us no clue as to why this should be
so, or what processes, relationships and perceptions produce this state
of affairs.
In her treatise on notions of causation in international relations Kurki
describes the underlying problem like this:
[T]here are countless topics in political science for which the quantitative com-
parison of many countries is simply inappropriate. This method of analysis
cannot ‘unpack’ important historical, political and sociological relationships
at a lower level of analysis. It cannot be used to analyze discrete moments of
political negotiation, consensus building, or the establishment of particular
political ‘pacts’ between elite groups and masses. It cannot be used to examine
different political strategies adopted by social movements, trade unions, revo-
lutionary movements, or other forms of collective action. It is not appropriate
for ‘process tracing’ in an effort to establish the links at the domestic level made
between actors and their propensity to reform political institutions. It cannot
map the inter-subjective meanings and different cultural understandings of
political concepts and practices, among other deeper processes of meaning
investigated through different methods of comparative analysis. (Landsman,
2008, p. 64)
176 Continuity and change in public policy and management
It will not have escaped the reader’s attention that most if not all of the
items on the above list are also issues with which our Anglo-Belgian study
has been centrally concerned. We have tried to ‘unpack’ relationships
(such as those between central government and the English police forces);
to analyse discrete moments (such as the launching of the UZ Leuven
project directly after de splitsing); to examine political strategies (such
as the increasing central control of National Health Service hospitals)
and to trace processes (such as the growth of professionalism among the
police). We have also pointed to the different cultural understandings of
the proper role of politics – for example with respect to local police forces
in Belgium as compared with England.
In fact strict adherence to the prescriptions of KKV would have made it
impossible to investigate many if not most of the issues we have covered.
More generally, ‘[I]f a sampling logic were applied to all types of research
. . . many important topics cannot be investigated’ (Agranoff and Radin,
1991, p. 205). For us, measures of suitable ‘variables’ did not exist or,
where they could be conjured up, they would not cover anything like
the whole period that we wanted to study. It is true that we could have
attempted to collect and build original databases ourselves, but prima
facie such an enterprise appeared unlikely to have much success. Even if it
had been possible, it would have been a mammoth undertaking, requiring
the investment of many person years of research time, for a most uncertain
outcome.
None of this should be taken as some root-and-branch opposition to
quantification. On the contrary, we have included statistical data in the
book wherever it existed and seemed relevant. However, that quantitative
material has always been set within a broader analytic narrative that has
gone far beyond dependent and independent variables. It has been used in
combination with qualitative observations to explore and redefine rather
than to test for acceptance or rejection.
Phenomena can only be understood within the context in which they are
studied; findings from one context cannot be generalized to another; neither
problems nor their solutions can be generalized from one setting to another.
(Guba and Lincoln, 1989, p. 45)
They also argue that their own kind of evaluation research has no special
legitimacy or status – it is ‘simply another construction to be taken into
account in the move towards consensus’ (p. 45). We see here some of the
characteristics referred to above: social reality is ‘constructed’ by each of us
and there is no special method or language of construction which is ‘better’
than any other. It therefore behoves the researcher to be modest about
his or her ‘findings’, and not to regard them as superior to the interpreta-
tions of policy practitioners or, indeed, everyday citizens. The evaluator,
in Guba and Lincoln’s vision, becomes a kind of negotiator, going round
all the stakeholders in a given situation and seeking to facilitate sufficient
inter-subjective agreement to serve as a basis for a consensus solution.
Some radical constructivists are happy to abandon all claims to objec-
tivity (or rather to deny that the idea of objectivity makes any sense).
Others, however, seek to hang on to that criterion, but do so by substan-
tially redefining it:
We trust that it is clear that our Anglo-Belgian comparison does not fall
into this camp. Ideas are certainly important to our analysis (the idea of
Flemish autonomy; the idea of neighbourhood policing) but so are mate-
rial factors that can only be reduced to ‘beliefs, traditions and dilemmas’
by a huge and quite artificial effort (factors such as limited finance, leaking
roofs, DNA testing and racially motivated riots). We want and need a
concept of causation that will in some way embrace these kinds of factors
as well as just ‘beliefs’. Of course the leaking roof or the budgetary shortfall
has to be ‘perceived’ and ‘believed in’ by the relevant actors, and in that
trivial sense may be reduced to beliefs, but to treat them primarily or exclu-
sively in that way strikes us as perverse. Nor are we content with Bevir and
Rhodes’s attempts to rescue objectivity. One does not have to go very far
back in time to find quite large communities where most people believed
that people with black (or yellow, or brown) skins were intrinsically less
intelligent than people with white skins. Or that society was arranged by
Reflections on doctrines of comparison 179
Our own paradigm falls between these poles and is usually labelled ‘critical
realism’. Critical realists hold that social structures and material factors
have a ‘real’ (ontological) existence, and that they can play a part in causa-
tion. Yet at the same time they are to a degree – but not to a radical degree
– ‘social constructivists’:
● Material causes. Social structures are real and have real effects.
Tenants are in a specific relation to their landlord and cannot just
choose to think or behave in some other way (unless they want to
lose their accommodation). Hospital chief executives have to stay
within or close to their budget ceilings (or find a quite exceptional
justification for not doing so), otherwise the institutional structure
(governments, supervisory boards) will ensure that they are disci-
plined or removed. The advent of forensic DNA testing leads to
old crimes being solved that could not previously be closed for lack
of acceptable evidence – the changing technology, although by no
means the ‘sole’ or ‘final’ cause, conditions what can be achieved at
a given moment. As one Chief Constable said in interview ‘no cars
were stolen in the nineteenth century’.
● Formal causes. Individuals are constrained and enabled by ideas,
rules, norms and discourses which surround and pre-exist them
and which are difficult to escape or avoid. Like material causes,
they condition actors rather than actively bringing about specific
effects. We heard, for example, that during the 1970s there was a
strong norm in Belgian federal politics to the effect that, if one of
the two main regions (Flanders, Wallonia) received major federal
support for a project, then the other region was entitled to some-
thing similar. There was thus a rough parity between financial
support for the new Flemish UZ Leuven, and support for the new
Reflections on doctrines of comparison 181
One does not have to follow precisely this typology to accept the general
point that critical realism embraces a more catholic and diverse concept of
causes than either the orthodox variables paradigm or the radical construc-
tivists (who are sometimes not sure if they have any time for ‘causes’ at all).
In another, influential formulation, Pawson and Tilley (1997) present
critical realism in a slightly different, though complementary way. They
stress that realist explanations of policy outcomes depend on the inter-
actions between specific mechanisms and particular contexts. So it is
misguided just to look for mechanisms which have worked (such neigh-
bourhood policing or closed-circuit television (CCTV) in car parks) –
mechanisms sensibly cannot be divorced from their contexts (which is a
problem for the variables paradigm, which tends to divorce variables from
their contexts). Furthermore, contexts are complex. Pawson and Tilley’s
182 Continuity and change in public policy and management
We have now set out two ‘polar’ paradigms, plus our preferred in-between
one (they are summarized in Table 9.1). In a moment we will move on to
the more specific issues of comparisons, case studies and narratives. For
clarity, however, we should insert a brief but very important caveat. It is
that in practice academics writing on public policy and administration are
constantly straying over the boundaries between these philosophically dis-
tinct perspectives. These boundaries are therefore porous. Interpretivists
will slip into talking of independent variables or the orthodox hardliners
will begin to interpret what the actors in their studies might be thinking.
Often this is done unconsciously, in the sense that the academics con-
cerned are focused on the topic of their study and reach for any insight
or piece of evidence that may be at hand, without worrying too much
about exactly how it has been generated and what its epistemological
underpinnings might be. At other times academics will quite explicitly and
consciously position themselves as occupying ‘mixed’ positions (indeed, it
could be argued that critical realism is itself a mixed position, insofar as it
insists on the importance of contextual interpretation and analysis whilst
at the same time searching for ways to get closer to the underlying reality
which it presumes is ‘just out there’). Thus it is not unusual, for example,
to find quite hard-nosed quantitative researchers emerging from their
multiple regression analyses in order to proclaim that ‘there are multiple
perspectives on social phenomena and . . . one’s own data are likely to
Reflections on doctrines of comparison 183
9.6 COMPARISONS
hand, as mild constructivists, critical realists are cautious about how far
across space and time it is plausible to make comparisons. ‘[I]nstitutions
and their effects can only be understood in particular spatio-temporal
contexts’ (Kay, 2006, p. 22). So there would be a reluctance, for example,
to compare hospital policy in Belgium with hospital policy in China. On
the other hand, realist theories, concepts and metaphors do have some
portability – they are not strictly confined to one spatio-temporal location.
What is important is to try to ensure that the respective contexts being
compared are carefully analysed and, if possible, are not wildly different.
It is in this sense a more holistic approach than the variables paradigm.
That is one reason why we chose two countries which were similar in
many socio-economic respects but which were obviously different in the
one important element of their political systems. The orthodox compara-
tivist would no doubt say that this was an MSSD if we wanted to make
socio-economic conditions our independent variable, but an MDSD if
we wanted to make the type of political system our independent variable.
Unfortunately for our methodological purity in this sense, we wanted to
look at both at the same time (as well as at a whole range of ‘outcomes’)
and we did not want to start off by defining them as variables incorporated
in specific hypotheses. We wanted to tell the stories as fully as we could,
and then look for patterns, not make up the story beforehand and then
check to see whether bits of what actually happened fitted the plot (or not).
It is interesting that a recent mainstream English comparative text like
Landsman (2008) does not even mention this possibility, despite the fact
that this is what many researchers in the fields of public policy and man-
agement are actually doing. Like KKV, Landsman sees small N basically
as a problem, not an advantage:
[C]omparative scholars will always face a trade-off between the scope of coun-
tries included in one study and the level of abstraction and strengths of the
inferences that result from the number of countries that are compared. Unlike
the global comparison of many countries in which sample sizes are maximized
for increasing variation in the variables of interest, comparing few countries
involves significant and intentional choices, any one of which may limit the
inferences made possible. The problem of selection bias looms large, the choice
of most similar and most different cases can appear at times arbitrary . . .
and the inclusion of negative cases, while laudable, may nonetheless not have
exhausted all cases that ought to be considered when analyzing particular out-
comes of interest. (Landsman, 2008, p. 81)
Now we can turn from general epistemology and ontology to the more
specific question of case studies. The present book consists of four case
studies at two different levels: national and local. A crude orthodox cri-
tique would be to say. ‘Only four cases, what can you prove with such a
small sample?’ But KKV themselves are considerably more sophisticated
than this. They are by no means opposed to small N case studies. Indeed,
they make an important point (with which we wholly agree) namely that:
Although case-study research rarely uses more than a handful of cases, the
total number of observations is generally immense. It is therefore essential to
distinguish between the number of cases and the number of observations. (King
et al., 1994, p. 52; see also Rueschemeyer, 2003)
before you get into the main part of the research study’ (Robson, 2002, p.
4). In flexible designs, although there is certainly plenty of planning, ‘much
less pre-specification takes place and the design evolves, develops and . . .
“unfolds” as the research proceeds’ (Robson, 2002, p. 5). We certainly
operated flexibly, finding and following up some issues that we were quite
unaware of when we first planned our research, and adding to our concep-
tual repertoire as we went along. However, ‘flexible’ does not mean ‘shape-
less’ or ‘anarchic’ – we spent quite a lot of time choosing our countries and
services and defining what kind of focus and data to go for. Furthermore,
at the micro-level we were reasonably meticulous with our interviews,
providing each person we approached with a summary of the project in
advance, then interviewing according to a carefully devised schedule of
questions, and finally sending the interviewees copies of our interview
records for them to correct or amplify (see the Appendix for details).
However, the amount of prespecification or flexibility is only one of the
restrictions which are applied by the orthodox school. For the main they
also see case studies as having a limited range of functions. First, they have
some value in testing existing hypotheses – a single well-chosen and well-
specified case (chosen and specified in relation to a defined population of
cases to which the relevant hypothesis is supposed to apply) can seriously
damage an unreliable hypothesis by showing that it is not confirmed.
(Note, though, that the situation is far more complicated with probabilis-
tic hypotheses, where one or two counter-cases do not dislodge the original
hypothesis; Mahoney, 2000. For probabilistic hypotheses, therefore, case
studies tend to lose even this function.) Second, they allow that in-depth
case studies may sometimes serve as useful preliminaries to hypothesis
formation. Thus, in a relatively unexplored field, a detailed case may lead
the researchers to see what they think may be a causal relationship, which
can then be incorporated in a formal hypothesis that can subsequently be
tested on a suitably defined large N population.
As for the radical constructivists, their attitude to case studies is arguably
paradoxical. On the whole they do not use the term, presumably because
it is tainted with ‘positivist’ and orthodox connotations. Yet a great deal
of their own work is what other social scientists would call ‘case studies’,
often quite small and local ones. At any event, the radical constructivists
provide us with no set of prescriptions specifically aimed at case studies.
Their methodological guidelines tend to be universal, in the sense that they
are held to apply to all scholarship, not just to certain subtypes.
In the territory between these two poles the past decade or so has wit-
nessed a surge of writing about case studies, the main outcome of which
can be said to be a much wider appreciation of the range of achievements
which can be hoped for from this form of research (Blatter and Blume,
188 Continuity and change in public policy and management
2007, 2008; George and Bennett, 2005; Gerring, 2007; Mahoney, 2000;
Rueschemeyer, 2003; Yin, 2003). Blatter and Blume (2008) have advanced
a particularly cogent ‘map’ of where case studies have made and can make
a contribution. They write of three main approaches to ‘case studying’,
each linked to a different epistemology.
First, ‘co-variance’, which looks to confirm or falsify correlations
between dependent and independent variables. This comes from the
‘variables paradigm’ we discussed above, and the function of the case is to
show whether and how the postulated independent variable(s) produces
the (dependent) outcome(s).
Second, ‘causal process tracing’, which aims to follow the mechanisms
that lead from prior conditions and interventions through to outcomes.
This holds out the potential to ‘get behind’ the statistical correlations of
the orthodox school and see into the ‘black box’ where this follows that
and then leads on to something different. It embraces a wider notion of
‘causes’ than the ‘variables paradigm’ would normally allow (Kurki, 2008,
Chapter 6). If a researcher is using a case or cases to trace processes then
the kinds of observations he or she will be pursuing and collecting will be
those which show quite detailed steps that lead from cause to consequence.
For example, we might be interested in the details of how the adoption of
the Resource Allocation Working Party (RAWP) formula for the distribu-
tion of financial resources between National Health Service (NHS) districts
across the UK meant that it was very hard for Brighton ever to receive a
large increment to its budget, because there were always other districts –
mainly geographically very distant ones – which were calculated as being
‘more deserving’. It is unlikely that the original architects of the RAWP
formula ever looked at the particular case of Brighton, but their principled
formula nevertheless had this consequence. Central government could
not make an exception of Brighton, however persuasive its pleas, because
once the whole issue was embedded in a centralized, national system, any
‘special deals’ would call forth demands and protests from the many other
districts that could argue that they too had a strong case for reasons that
lay outside the formula.
Third, ‘congruence analysis’, where the researcher is comparing, on the
one hand, a set of expectations that can be deduced from theories and con-
cepts with, on the other, whatever empirical findings emerge from the case
or cases. This is not the same as the formal hypothesis testing of the vari-
ables paradigm because it is undertaken with what Robson (2002) would
no doubt term greater ‘flexibility’. Here theory is treated as an interpretive
framework (mild constructivism) but not as one which totally or almost
entirely determines empirical ‘findings’ (radical constructivism). The case
study is at its most powerful when a plurality of theoretical expectations
Reflections on doctrines of comparison 189
can be tested against it. When used in this mode, the most valuable obser-
vations are not those which would best enable process tracing, but rather
those which help the researcher to discriminate between two or more
competing theories. In our case we found the theoretical expectation that
the fragmented and consensual Belgian political system would be slower
and less decisive confirmed in the case of police policy at the national level,
but disconfirmed as far as hospital policy at the local level was concerned.
This usefully drew attention to another dimension of the contrast between
the two different political systems – that even if the fragmented and con-
sensualist Belgian system was slower-moving overall, its unpredictability
and lack of national ‘standardization’ meant that local breakthroughs
and projects (such as the new UZ Leuven) stood a better chance than
in the grindingly bureaucratic NHS. There were perhaps more ‘interior
windows of opportunity’ than in a more majoritarian and centralized
system. However, in the light of Blatter and Blume, we could undoubtedly
have done more to try out a wider range of rival theoretical models than
we actually did.
We therefore see the Blatter and Blume reappraisal of case studies as
being of considerable relevance to the analyses conducted in this book
(which include both some elements of process tracing and some congru-
ence analysis). They show that a wider appreciation of the nature of theory
coupled with a more plural concept of causation makes the small N prob-
lems identified by the orthodox look less important and less constraining.
Whilst our account may have been ‘critical realist’ and ‘mildly constructiv-
ist’ a large part of it could also be described as ‘narrative’. We describe and
explain by telling stories and, because these stories cover 40 years, we can
begin to think of them as historical narratives. Histories, of course, can
take many forms – some can be theory driven and others can be virtually
theory-free, at least in any overt sense – but most involve narrative (Pollitt,
2008, pp. 32–40). In almost every case the sequence of events is deemed
important. Some later things can happen only because some earlier things
have already happened. Diachronic comparisons reveal arrows as well
as cycles (Pollitt, 2008, pp. 51–9). In our study, for example, the rapid
construction of the new UZ Leuven on the Gasthuisberg would have been
unlikely unless it had been immediately preceded by the momentous politi-
cal events of de splitsing. And the considerable infrastructural investment
in the RSCH site in the 2000s would have been unlikely if the New Labour
government had not previously reshaped the Public Finance Initiative
190 Continuity and change in public policy and management
This would result in ‘a story, Jim, but not as we know it’. It would be a
highly stylized, preformed plot, with a carefully preselected list of char-
acters. It would be vigorously opposed by many mainstream historians,
but even more by the radical constructivists, who want all narratives to
be ‘decentred’, not told from one point of view with one hypothesis to be
tested, but to be exploratory and conjectural (Bevir and Rhodes, 2006)
We reject here the notion that narratives should be conceived as ‘testing’ the
model, on the grounds that to do so would inevitably render the narrative a
‘just so’ story where features of the world that are essential and causal in this
Reflections on doctrines of comparison 191
context are ignored because they do not have, nor could they have, a place
in the general model because of the irreducible complexity that characterizes
policy processes. (Kay, 2006, p. 63)
In historical narratives, theoretical models are used but they are local or con-
textual, and sometimes limited to one specific, temporally distinct event within
the narrative. Theory is always subordinate to the evidence. The burden of the
narrative is to weigh competing models, concepts or metaphors and show that
one is the most appropriate in view of the evidence. (Kay, 2006, pp. 61–2)
This means, of course, that our narratives often have loose ends, and
that they seldom confirm or falsify in orthodox, binary fashion. Instead
they are used in an analytic dialogue with the evidence, strengthening this
interpretation a little and weakening that one. They present a synthesized
assessment of the relative appropriateness of a series of explanations for
this unique body of evidence rather than the (dis)confirmation of the
applicability of a general hypothesis to a theoretical instance.
and epistemological marsh before they could get started on Belgium and
England.)
Certainly our study has not been a KKV-style orthodox ‘scientific’
study of observable variables. In particular it has not been shaped around
the formal testing of one or more defined hypotheses. And although it
has used plenty of statistics these have played a supportive rather than
a decisive role: it has not been first and foremost a quantitative analysis.
The first tasks we set ourselves (Chapter 1, section 1.1) were in fact those
of description rather than explanation. They were to:
It was clear from the outset, however, that description was neither a
simple nor a theory-free task. Our own attempts necessarily employed
concepts and categories drawn directly from theory – most notably the
distinction between a majoritarian and a consensus political regime, but
also concepts of centralization and decentralization, managerialism and so
on. We have a sneaking sympathy with Bruno Latour’s point that ‘if your
description needs an explanation, it’s not a good description’ (although
we would not go along with much of the rest of Latour’s ‘Actor Network
Theory’) (Latour, 2007, p. 147).
Theory has therefore figured as an input to description (as is often said,
there are no wholly theory-free facts or, to put it another way, no truly
naive descriptions). But there has also been a theoretical output. This has
mainly come in an inductive manner, not through the kind of hypothetico-
deductive processes that are pursued by the orthodox ‘variables paradigm’
type of social science. Various theoretical insights and possibilities have
thus emerged from our historical analyses, and can be tested by further
research. Just five may be offered here as illustrations:
195
196 Continuity and change in public policy and management
In 2004 one of us (Pollitt) was awarded the Hans Sigrist International Prize
in recognition of his previous comparative public management research.
The Hans Sigrist Foundation is connected to the University of Berne,
Switzerland. The prize money paid for most of the research that was
carried out in Brighton. At the same time, the organizational support and
base for this research was provided by the Institute of Public Management
at the Katholieke Universiteit Leuven (KUL). KUL awarded Pollitt a
Senior Research Fellowship from 1 September 2005 until 31 August 2006,
and from October 2006 he became a tenured professor there. Bouckaert
was Director of the Institute throughout the period of the project.
In Brighton Pollitt made arrangements to work with Professors Sue
Balloch and Michael Hill of the Health and Social Research Policy Centre
and with Professor Peter Squires, Professor of Criminology – all at the
University of Brighton. He remains warmly grateful to all of them for their
generous advice and support, and for the offer of a working base at the
University’s Falmer site.
References
Abbott, A. (1997) ‘Of time and space: the contemporary relevance of the
Chicago School’, Social Forces 75: 4, pp. 1149–82.
Adcock, R. (2005) ‘Generalization in comparative and historical social
science: the difference that interpretivism makes’, in D. Yanow and
P. Schwartz-Shea (eds) (2005) Interpretation and Method: Empirical
Research Methods and the Interpretive Turn, Armonk, NY, M.E.
Sharpe, pp. 50–66.
Agranoff, R. and B. Radin (1991) ‘The comparative case study approach
in public administration’, Research in Public Administration 1, pp.
203–31.
Allen, D. (1979) Hospital Planning: The Development of the 1962 Hospital
Plan, Tunbridge Wells, Pitman Medical.
Andrews, R., G. Boyne and R. Walker (2008) ‘Reconstructing empirical
public administration: Lutonism or scientific realism?’ Administration
and Society 40:3, pp. 324–30.
Arthur, B. (1994) Increasing Returns and Path Dependence in the Economy,
Ann Arbor, MI, University of Michigan Press.
Association of Chief Police Officers (1990) Setting the Standards for
Policing: Meeting Community Expectations, Strategic Policy Document,
London, ACPO.
Audit Commission (1990) Effective Policing: Performance Review in Police
Forces, London, HMSO.
Audit Commission (1993) Helping with Enquiries: Tackling Crime
Effectively, London, HMSO.
Audit Commission (1996) Streetwise: Effective Police Patrol, London,
HMSO.
Bache, I. and M. Flinders (eds) (2004) Multi-level Governance, Oxford,
Oxford University Press.
Ballantine, J., J. Forker and M. Greenwood (2008) ‘The governance of
CEO incentives in English NHS hospital trusts’, Financial Accountability
and Management 24:4, pp. 385–410.
Barclay, G. and C. Tavares (2003) International Comparisons of Criminal
Justice Statistics 2001, issue 12/03, London, Home Office paper.
Baumgartner, F. and B. Jones (1993) Agendas and Instability in American
Politics, Chicago, IL, University of Chicago Press.
198
References 199
Bulletin (1991) ‘MP challenges single-site hospital plan’, No. 246, April, p. 2.
Burrell, G. (1997) Pandemonium: Towards a Retro-organization Theory,
London, Sage.
Büthe, T. (2002) ‘Taking temporality seriously: modeling history and the
use of narratives as evidence’, American Political Science Review 96:3,
pp. 481–93.
Butler, Y. (2000) ‘Managing the future: a chief constable’s view’, in F.
Leishman, B. Loveday and S. Savage (eds) Core issues in policing, 2nd
edition, Harlow, Pearson Education, pp. 305–20.
Cabinet Office (2008) Excellence and Fairness: Achieving World Class
Public Services, London, Cabinet Office.
Canadian Institute for Advanced Research (1997) ‘CIAR estimated health
impacts of determinants of health on population health status’, http://
www.hc-sc.gc.ca/hppb/regions/ab-nwt/pdf/pop-health_e.ppt.
Carter, C. (2008) ‘Identifying causality in public institutional change:
the adaptation of the National Assembly for Wales to the European
Union’, Public Administration 86:2, pp. 345–61.
Castles, F. and D. Mitchell (1993) ‘Worlds of welfare and families of
nations’, in F. Castles (ed.) Families of Nations: Patterns of Public
Policy in Western Democracies, Aldershot, Dartmouth, pp. 93–128.
Colebatch, H. (1998) Policy, Buckingham, Open University Press.
Collier, P. (2006) ‘In search of purpose and priorities: police perform-
ance indicators 1992–2004’ Public Money and Management 26:3, pp.
165–72.
Cumberledge, J. (1985) quoted in Brighton Health Bulletin, No. 188,
January, p. 3.
Davies, A. and R. Thomas (2008) ‘Dixon of Dock Green got shot! Policing
identity work and organizational change’, Public Administration 86:3,
pp. 627–42.
Delarue, D. (2001) Algemeen politiemanagement: managementtechnieken
et concepten voor de politie (General police management: management
techniques and concepts for the police), Brussels, Politiea.
Department of Health, Welsh Office, Scottish Home and Health
Department, and Northern Ireland Office (1989) Working for Patients,
Cm. 555, London, HMSO.
Deschouwer, K. (2006) ‘And the peace goes on? Consociational democ-
racy and Belgian politics in the twenty-first century’, West European
Politics, 29:5, pp. 895–911.
De Somer, P. (1985) Speech at Plechtige inhuidiging van het Universitaire
Ziekenhuis op de gasthuisberg te Leuven, Leuven, Katholieke
Universiteit Leuven, pp. 3–9 (inauguration ceremony for the Leuven
University Hospital).
References 201
Sussex Police (1992) Sussex Police Annual Report 1992, Lewes, Sussex
Police Printing Department.
Sussex Police (1993) Sussex Police Annual Report 1993, Lewes, Sussex
Police Printing Department.
Sussex Police (1995) Sussex Police Annual Report 1994/5, Lewes, Sussex
Police Printing Department.
Sussex Police Authority (2003) Joint Annual Report of the Sussex
Police Authority and the Chief Constable of the Sussex Police 2002/3,
Chichester, Sussex Police Authority.
Sussex Police Authority (2004a) Joint Annual Report of the Sussex
Police Authority and the Chief Constable of the Sussex Police 2003/4,
Chichester, Sussex Police Authority.
Sussex Police Authority (2004b) Local Policing Plan for Sussex, 2004–2005,
Lewes, Sussex Police Authority.
Sussex Police Authority (2005) Joint Annual Report of the Sussex
Police Authority and the Chief Constable of the Sussex Police 2004/5,
Chichester, Sussex Police Authority.
Tange, C. (2004) ‘Evaluating the police: apolitical, professional and sci-
entific issue – 25 years of experience evaluating the police in Belgium’,
European Journal of Crime, Criminal Law and Criminal Justice 12:3, pp.
232–50.
Team Consult (1988) Les services de police en Belgique: rapport au Ministre
de l’Intérieur, PGR, Brussels, INBEL.
Team Consult (1995) ‘Evaluatie van de werking van het Korps van
Leuven’ (Presentatie Nr. 4, 20 Januari) (‘Evaluation of the workings of
the Leuven police corps’), unpublished report.
Test Gezondheid (2005) ‘De reputatie van ziekenhuizen: artsen aan het
woord’, No. 69, Oktober/November, pp. 12–16.
The Argus (1995) ‘Down but not out!’, 12 April, p. 1.
The Argus (2005) ‘Sussex NHS vows to improve’, 28 July, p. 1, http://
archive.theargus.co.uk/2005/7/28/201527.html, accessed 1 June 2006.
Thelen, K. (2003) ‘Insights from comparative historical analysis’, in J.
Mahoney and D. Rueschemeyer (eds) Comparative Historical Analysis
in the Social Sciences, Cambridge, Cambridge University Press, pp.
208–40.
Thompson, J. and H. Rainey (2003) ‘Modernizing human resource man-
agement in the federal government: the IRS Model’, Washington, DC,
IBM Endowment for the Business of Government.
Tilley, C. (2006) ‘Why and how history matters’, in R. Goodin and C.
Tilley (eds) The Oxford Handbook of Contextual Political Analysis,
Oxford, Oxford University Press, pp. 417–37.
Tilley, N. (2008) ‘Community policing’, in T. Newburn and P. Neyroud
References 209
211
212 Continuity and change in public policy and management
elite networks, Belgium and England fiscal crisis in 1976 and 1980s, National
139 Health Service 39
emergency calls, rapid rise, 1980 103 Fisher Report on police
England rights violation of teenage boys
Area Health Authorities 39 charged with murder 52
public policymaking and Flemings and Walloons, conflict 24,
management 1 83, 92
no national police force 101 Flemish autonomy 153, 155
similarities to Belgium 10‒11 Flemish Brabant province,
English hospitals 32‒4 Leuven, administrative capital 116
market-type mechanisms, increased Flemish Christian Democratic elite
use 72 interaction 94, 96, 194
English hospitals policy, 1965‒2005 football hooligans, Heysel Stadium,
rationalization drive 37‒43 Brussels, 1985
ethnic equality in police force 116 deaths and injuries 62
ethnic minorities, urban disorder and foot patrol, efforts to direct more
police riots, London, 1980s 28, 54 officers, 103‒4
Euro 2000 Football finals, Belgium Force Crime Strategy, 1994 109
international security concerns 65 form filling by police 31
Eurobarometer, satisfaction with Foundation Trust (FT) 42
democracy 163 Francophone students, Thursday riots
Eurobarometer, trust in Civil Service, in Leuven, 1975 121
Belgium/Britain 162‒63, 166 Francophone (Walloon) university,
European Council of Ministers 54 Louvain-la-Neuve, 83, 93
European Observatory on Health Care funding needs, Leuven 125
Systems, 2000 35, 43 funding of Sussex police, 2002‒03
European single market 112
European Union membership,
marginal influence 138 Gasthuisberg plan 194
government welfare reductions 165 Gatwick airport, Sussex, special police
Europol Organization, Maastricht unit 102
Treaty 54 Gendarmerie
expansion and specialization, Belgian district staff merger with Leuven
police 61 118
expenditure rises for hospital municipal police differences 129‒30
pharmaceuticals, Belgium, 1996 29 gender equality in police force 116
general practitioners in medicine 38
federal-level hospital policy, Belgium geriatric ward, indefensible nursing
92 practices 100
Federal Security Plan, Belgium, 2000 TV documentary on Brighton
65 hospitals 90
fee-for-service payment of providers, governance networks, different in
Belgium 43 England 139
financial mechanisms, NHS capital government influence
budget 156 Leuven and Sussex, comparison
financial support for Brighton 130‒31
hospitals 87 government initiatives, impact of, on
financial targets 89 police 113
fire service, and origin of police, government intervention in hospital
Leuven 120 system 30, 31, 48‒9
Index 215
Police Reform Acts, 1998, 2002 27, 56, Prosecution of Offences Act, 1985 54,
59, 64‒5, 151 106
police services 1965‒85, comparison, public health insurance, compulsory in
Belgium /England 28 Belgium 34‒5, 43
high-tech operations 31 public order breakdowns, England
reduction in Belgium 28 31
police specialization 136 public policy change, levels of
police statistics, selected, England and examples in National Health Service
Belgium 76 (NHS) and police 7
police training in Leuven 120‒21 public policymaking and management
police training in technology 31 Belgium and England 192
Policing Improvement Agency, 2004 72 public reporting of crimes, variable 58
Policing in Sussex, 1990 Public Service Agreements (PSAs),
public opinion survey, University of England 30, 57
Sheffield team 108 public trust decline in police and
policing, traditional ‘beat’, prevention criminal justice system 136
of, by other duties 105‒6
policy activism for police, debatable Quality of Service (QOS) initiative 54
good 4, 75
policy aspects 3‒4, 7 radical conservatism 145
policy instruments, changes in levels 71 radical constructivism 176‒9, 183, 187,
policymaking as ‘bright ideas’, 191
England 40‒41 rainbow coalition in Belgium, 1999
policymaking for English hospital liberals, socialists, environmentalists
system 65
government intervention increasing rationing of NHS investment 147‒8
43 recorded crime, tables of, England and
political control over police, Sussex, Wales, 1965‒2002
tightening 131 statistics, unreliability of 59‒60
political systems of England and reforms, intensity of, Belgium and
Belgium 10‒11, 16, 23‒6, 137 England
politicians, local, role of, in Leuven comparisons over time 71
130 Regional Directorates of Health and
politicization of public service 27 Social Care, 2001 42
popular culture, hospitals and police 10 Regional Health Authorities, England
powers, devolvement of, in Belgium 38, 40, 89
to Wallonia, Flanders, Brussels 24 Regional Offices of NHS 42
Primary Care Trusts, ‘independent regulation by central government,
treatment centres’ 42 increase in England 60
Princess Royal Hospital, Haywards resignation of Ministers of Justice,
Heath 88, 90 Belgium 64
prisoner transportation contract 55 resistance to centralised reorganisation,
Prison Policy, Belgium, 2000 65 England 168
private health care expenditure Resource Allocation Working Party
Belgium in 2000, lowest rate 47 (RAWP), 1976 188
private security industry 66, 73 allocation of NHS revenue funds 89
probation services 73 resource allocation, health
professionalization of local police 60, authorities 39
72 respect for authority, loss of, in
proportional representation system 161 population 116
Index 221